• Location: MIDDLESEX 
  • Founded:
  • Website:


  • Page 1

    Cisco 2008 Annual Security Report Highlighting global security threats and trends

  • Page 2

    Contents Title The Cisco Duisi ® Annualeetuer tate molobor Security sentReport provides anod am quametuero overview magna of the combined feugait in ullamet security nos ad teintelligence minissed tatoflorperillam, the entire Cisco organization. quatetum Theciliquisit aliquis report encompasses threat vel dolenim zzril information iustrud estrud and trends duisi collected eliquam alit between January iriliquis nullaore and October feugue vulpute2008,voloreand providesnullaor magnat a snapshot of the state at ad eugait of security velessit for core ad elit lam, that period. faccum The report augiati also provides onsequatem recommendations iliquisit from Ciscoquipis venibh eu faci esequatem security ercilit loreexperts con verci andblampredictions nulputem of quis how nulla identified trends will aliquat. continue to unfold in 2009. Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 3

    2 Introduction 16 Data Loss TCP Stack Table Implementation Vulnerability Online Threats Data Loss Issues on the Regulatory Radar Networking Equipment Vulnerabilities The Web The Limitations of Compliance Virtualization Vulnerabilities Malware Recycling Risks Encryption Vulnerabilities Botnets Identity Theft Operating System Vulnerabilities Spam Targeting the Masses Vulnerabilities in Databases and Data Loss Rethinking Identity Management Office Productivity Applications Insider Threats Mobile Device Vulnerabilities Vulnerabilities 20 The Human Factor Top Security Concerns of 2008 Human Nature Invites Risk 38 Geopolitical and Remote Working, Social Networking: Political Conflicts Opportunities and Risks 6 Online Security Risks From Conflicts to Cybercommands Using Social Networking and Web 2.0 Sites and Trends for Online and Offline Crime Web Trends 41 Countering Internet Compromising Legitimate Websites 24 Insider Threats Security Threats Popular Methods of Compromising DNSSEC Legitimate Websites Financial Crisis May Heighten Insider Risk Industry and Government Initiatives Malware Trends Enabling Technologies Mobile Phone Malware: Growing Profit Centers 26 Issues of Trust Making Security Easier to Find The “Shadow” Internet Economy New Tactics Erode Trust Asprox: Transforming an Old Trojan Privacy and Trust Violations 44 Conclusion and Key Botnet Trends Recommendations The Importance of Social Engineering 30 Vulnerabilities Putting IT on the Front Lines New President, New Malware Web Vulnerabilities Key Recommendations Beijing Olympics Fake Ticketing Scams ActiveX Vulnerabilities Top Trends to Expect in 2009 Spam and Phishing Trends DNS Vulnerabilities A Holistic Approach to Security Spear-Phishing Examples Recent Attacks Using DNS Cache Poisoning Email Reputation Hijacking The Importance of DNS DNS Cache Poisoning All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 1

  • Page 4

    Introduction Title There was an enormous amount of activity related to data and online security during the past year. Although no single, overwhelming attack—such as the spread of Melissa, Slammer, or Storm malware in previous years—turned into the signature security event of 2008, the need for increased security protection and continued vigilance remains. 2 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 5

    Compared to previous years, online criminals are Online Criminal Ecosystem becoming even more sophisticated and effective, employing a greater number of relatively smaller, more Legitimate users targeted campaigns to gain access to sensitive data. Human nature—in the forms of insider threats, susceptibility to social engineering, and carelessness that leads to inadvertent data loss—continues to be a major factor in countless security incidents. And the increasing use at many organizations of technologies designed to increase collaboration and productivity (such as mobile devices, virtualization, cloud computing, and other Web-based Legitimate users visiting subverted sites, invisibly downloading malware tools and Web 2.0 applications) is stretching the edges of corporate networks, potentially increasing security risks. Many different entry points or “threat vectors” are used to compromise the security of individuals and organizations. Criminals creating malware and hacking legitimate websites For example, threats can be aimed at mobile devices and insecure hardware; at weaknesses in operating systems, office productivity applications, and encryption tools; and The Web changes an infected computer’s activities—for example, at numerous other vectors. In the online threat arena, the entire Web ecosystem causing it to connect to the Internet or install additional comes into play. Online criminals continue to create malware without the user’s knowledge. Other malware Online Threats malicious websites—carefully designing them to look works to find sensitive information, such as user passwords In terms of quantity and pervasiveness, the most alluring and legitimate—to obtain sensitive personal and credit card numbers, on a computer or network, and significant security threats in 2008 involved an online information or distribute malware to site visitors. They sends that information “home” to online criminals. In addition, component. These online threats continue to grow in hack legitimate websites from trusted organizations, an increasing amount of malware is being developed scope and number, and should remain a top concern such as news media or large retailers, to cause those and sold. for security professionals. sites to invisibly distribute malware to visitors; they also Many of these online threats combine the following create or subvert existing Web applications and plug-ins Botnets closely related elements: for the same purpose. In addition, in the core underlying The core mission for much of today’s malware is to infiltrate • The World Wide Web infrastructure of the Internet, weaknesses have been a computer and make it part of a botnet. Botnets consist of • Malware exposed that could let online criminals divert thousands of thousands of malware-compromised computers (botnet • Botnets unsuspecting Internet users at once to malicious websites. nodes or “zombies”), and they have become the cornerstone • Spam of large-scale online criminal activity. The people controlling Malware botnets can rent out the processing power and bandwidth Although far from the only method, the Web has become of these subverted computers to others, or use it them- the primary means of infecting computers with malicious selves to send out massive amounts of spam, attack software. Most modern “malware” is designed to help websites, or engage in other nefarious behavior. someone gain control over a computer, communications device, or network. Some malware directly influences or All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 3

  • Page 6

    Spam by Originating Country for 2008 Many spammers still blast out “mass-mailing” spam to • The website downloads malware onto the site millions of untargeted recipients per campaign; many visitor’s computer to gain control over it. Originating Country Percentage of Global Spam anti-spam products work on filtering out these types • The compromised computer becomes part of a USA 17.2% of messages. But for more sophisticated “phishing” botnet, and starts sending out spam. Turkey 9.2% Russia 8.0% spam—which is designed to elicit personal or financial Canada 4.7% information—smaller, more targeted campaigns are Data Loss Brazil 4.1% becoming the norm. Data loss often occurs through the loss or theft of India 3.5% Spammers continue to improve the design and effective- equipment such as laptops or removable storage media, Poland 3.4% Korea 3.3% ness of their messages. They’re using highly topical or when a computer or network is infiltrated to steal Germany 2.9% subject lines, far more legitimate-looking and professional- sensitive data or intellectual property. Fewer organizations United Kingdom 2.9% and individuals may be affected by data loss incidents sounding content, and other techniques that make certain Thailand 2.8% types of spam hard to resist for normally wary recipients— than by online threats, but the impact of these events can Spain 2.8% Italy 2.4% and easier to slip by anti-spam solutions. be devastating. Argentina 2.1% For organizations that experience data loss, reputations To actually send out their spam messages, online criminals Columbia 2.1% rarely use computers in their physical possession, instead and trust can be damaged or destroyed, while financial France 2.0% Other 26.7% renting or building botnets to do the mailing for them. This consequences such as stock-price drops, lawsuits, and completes an elegant cycle, in which: compensatory damages to affected individuals can run into millions or even billions of dollars. For individuals, the Spam • Botnet nodes send out spam. consequences of a data loss incident that compromises Spam, or unsolicited email, is one of the most pervasive • Spam recipients get an email message that lures highly personal information, such as Social Security Internet threats, affecting nearly every Internet user and them to a malicious website. numbers or financial details, can negatively affect their organization in the world. Different types of spam include: lives and finances for years. • Email messages promoting items such as pharmaceuticals, printer cartridges or corporate Average Daily 200 equity instruments for sale. Spam Volume 180 Daily Spam Volume (in billions) • Email messages with an attached file that contains 160 malware. 140 • “Phishing” emails that lure recipients into providing 120 personal information via a return email or by filling 100 out forms on a website. 80 • Email messages that include URLs and attempt to 60 convince recipients to visit seemingly trustworthy 40 websites that actually distribute malware. 20 0 May-07 May-08 Mar-07 Mar-08 Jun-07 Jun-08 Apr-07 Apr-08 Aug-07 Oct-07 Aug-08 Oct-08 Jan-07 Jan-08 Dec-07 Jul-07 Nov-07 Jul-08 Sep-07 Sep-08 Feb-07 Feb-08 Daily spam volumes have nearly doubled in 2008 relative to 2007. Month All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 4 Cisco 2008 Annual Security Report

  • Page 7

    Legislation and industry initiatives focused on making data on networks more secure and informing parties affected Top Security Concerns of 2008 of data breaches are increasing. Many organizations are working to better enforce their existing acceptable use Threats and criminals are becoming Criminals are exploiting vulner- “Invisible threats” (such as hard- policies around sensitive data. Yet compliance with such faster, smarter, and more covert. abilities along the entire Web to-detect infections of legitimate policies and initiatives is not a guarantee of safety, as the ecosystem to gain control websites) are making common • Specialization and innovation in the growth and evolution of data loss threat factors are likely of computers and networks. sense and many traditional security online crime economy continues. to outpace the initiatives or legislation addressing them. solutions ineffective. • Botnet infestations remain • Attacks are increasingly targeted to common and dangerous. Insider Threats help maximize their effectiveness. Loss of data and intellectual • Known vulnerabilities are going Sometimes, the people responsible for data loss and • Many types of reputation hijacking property are continual challenges. unpatched and existing security other security incidents are insiders, including current (attacks that exploit users’ trust in policies are being ignored. • Data loss is often caused by someone’s reputation) are gaining or former employees who want to cause trouble or are exploiting vulnerabilities in in prevalence and popularity. • Widespread use of Web-based simply looking for personal gain. This type of threat can technology and human nature. collaborative technologies in the be especially grave, as insiders know the weaknesses in • Blended threats that combine workplace brings added risks as • A company’s reputation, trust and an organization’s security and how best to exploit them email and websites and use social well as greater productivity. finances can be affected. engineering techniques are now to steal data or money, or even hold assets for ransom. more common than ever. • Risk vectors include online threats, In today’s uncertain economy, in which more employees mobile devices, and insiders. may lose their jobs or become dissatisfied with their work situation, and in which less budget may be available to address security concerns, insider threats —and the likelihood of their success—are of increasing concern. Other vulnerabilities can be exploited as well, including Vulnerabilities (among others) weaknesses in office productivity In addition to taking advantage of aspects of human nature applications, operating systems, mobile device tech- (such as curiosity, trust, and carelessness), criminals are nologies, networking equipment, virtualization tools, and getting access to computers and networks by exploiting encryption technologies. However, vendors of affected weaknesses in technologies, software, and systems. products are now often disclosing vulnerabilities—and releasing patches at the same time—to mitigate the effects In 2008, vulnerabilities in the entire Web ecosystem— of the vulnerability, making staying up-to-date on patches browsers; helper objects, media players, and plug-ins more important than ever. running in those browsers; Web server and application software; and core parts of the underlying infrastructure of the Web—were exploited to gain control of computers, networks, and data. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 5

  • Page 8

    Online Security Risks and Trends Online security threats continued their growth in 2008. Online criminals combined spam, phishing, botnets, malware, and malicious or compromised websites to create highly effective blended threats that use multiple online vectors to defraud and compromise the security of Internet users. 6 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 9

    The Web itself is a primary mechanism for distributing As the possibilities and popularity of the Web have grown, More and more of these malicious websites involve malware that lets someone gain control over computers so has its use as a threat vector. “in-depth” attacks, where several different types of and networks. Many of these computers are then turned Originally, malicious software was distributed via floppy exploits that work on different weaknesses (in different into nodes in a botnet, where they engage in a variety of drives and macros in infected office documents, then browsers, plug-ins, and operating systems) are hosted on activities, usually without the computer user ever being via network worms such as Slammer, followed by an the same website. This increases the chance that each aware that this is happening. enormous rise in distribution via email. Today, a vast website visitor will display a weakness—and one is all These activities can include sending massive volumes of quantity of malware is downloaded from websites. that’s needed—that the exploits can take advantage of to spam—designed to either lure more victims to websites Criminals exploit vulnerabilities throughout the entire Web download malware to the computer. where they’ll download malware, or to obtain personal ecosystem to gain control of computers and networks. Compromising Legitimate Websites information—hosting malicious websites or infecting (For more specific examples and information, see the A method of propagating malware that reached new levels legitimate websites, and helping to overwhelm websites Vulnerabilities section later in this report.) of popularity in 2008 is compromising legitimate websites or computer networks with distributed denial of service These malware infections often happen without any user to make them hubs for malware distribution. In April 2008 (DDoS) attacks. intervention or awareness in what is known as a “drive-by alone, thousands of websites were compromised and download.” Someone visits a malicious or infected website tried to infect site visitors with malware. Web Trends hosting exploits that look for weaknesses in the site visitor’s Fifteen years ago, barely anyone knew what the World Causing a trusted legitimate website to host exploits browser or computer system. If the exploits detect a usable Wide Web was. Today, Google is one of the top trusted or serve up malicious code is an effective way to infect weakness, they start trying to download malware to the brands in the world, and people extensively use the Web computers with malware. Site visitors have no hesitation computer. This can all happen quietly in the background, for everyday activities such as communication, research, about the trustworthiness of the site; it is a legitimate site without the site visitor ever clicking on a link in the infected shopping, and financial matters. they visit for content or transactions on a regular basis. page or finding out what’s going on. Internet security applications that depend on URL or IP Unlike its earliest predecessors, the modern Web browser address filtering also trust the website’s legitimacy. provides an amazing, highly interactive experience. Flash animation ads play within a webpage; audio streams on the Vulnerability and Threat Categories for 2008 And when legitimate websites are infected, specific websites of bands or online radio stations let site visitors user groups can be targeted with great precision—for listen to music; videos play automatically; social networking buffer overflow example, infecting sites aimed at students, online gamers, denial of service sites and widgets integrate contact information, photos, or or business users, with the latter potentially providing arbitrary code execution data from a person’s blog with their social networking page; cross-site scripting channels into business networks through compromised and Adobe PDF documents render seamlessly within privilege escalation workplace computers. information disclosure the browser. software fault Cisco data shows that exploited websites are currently This is all possible because the Web browser uses plug- directory traversal responsible for more than 87 percent of all Web-based backdoor trojan ins, media players, browser helper objects, and tools like threats. And according to security audit provider White unauthorized access ActiveX controls and JavaScript commands to activate spoofing Hat Security, more than 79 percent of the websites different types of objects on a webpage. Underlying Web format string hosting malicious code are legitimate websites that have worm applications such as content management systems show been compromised. Nine out of any 10 websites may be security solution weakness the right content at the right time, while forums and wikis vulnerable to attack: Seven out of 10 are susceptible to let site visitors quickly post to and modify webpages. cross-site scripting (XSS) exploits, and one in every five In 2008, vulnerability and threat activity was dominated by buffer may be vulnerable to SQL injection attacks. overflows and denials of service, with arbitrary code execution being the next most prominent category. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 7

  • Page 10

    Popular Methods of Compromising In one notorious case in September 2008, online criminals compromised hundreds of pages on the BusinessWeek.com Legitimate Websites website with a SQL injection attack. As one of the top 1000 visited sites on the Web, BusinessWeek.com enjoys a iFrame exploits. Both XSS and SQL injection exploits Cross-site scripting (XSS). A flaw within Web appli- high degree of trust from Web users. Naturally, that makes commonly use iFrames as a vehicle for delivering malicious cations that lets ill-intentioned users of vulnerable websites it extraordinarily attractive as a site from which to serve code, which can install malware on a computer without or owners of malicious websites send malicious code malware to unsuspecting visitors. the user’s knowledge. An iFrame, or inline frame HTML to the browsers of unsuspecting users. These attacks tag, can allow the embedding of compromised Web code are frequently executed using HTML image and frame Malware Trends from another Web server into a separate HTML document. elements (<img>, <frame>, <iframe>) and JavaScript. A vast amount of online crime and profit is enabled by Common applications of this method include setting the Cross-site request forgery (CSRF or XSRF). An exploit control of personal computers. The malware, or malicious size of the iFrame to zero and simply passing malicious in which an attacker uses the knowledge that the victim is software that infects these computers and turns them into code through the host site without the knowledge of the site visitor or the Web host. currently engaged in a browser session on one website to botnet nodes, is the first step. forge instructions ostensibly from the victim on another site 2008 was another banner year for Web-based malware. SQL injection. Exploits a security vulnerability in the where the user is persistently or currently authenticated. Online criminals continued employing the Web-based database layer of widely used Web applications and For example, the attacker and victim are both online in a servers. Recently, hackers have used Structured Query distribution techniques that worked so well for them Web forum, and the attacker is able to steal the victim’s Language (SQL) exploits to include malware or invisible links in 2007, and kept refining them further for greater authentication to make purchases at an e-commerce site to malware-hosting sites on legitimate websites. They did that the attacker knows a) is frequented by the victim, effectiveness and profit. this by taking advantage of website developers not properly and b) does not require re-authentication before finalizing Of the malware distributed via the Web, a large portion sanitizing data transmitted in user input fields (such as forms purchases. consisted of Trojans, designed to seem innocuous, and user logins) on webpages that use SQL. Thousands of invisible, or attractive before being installed. Rootkits, websites using Microsoft ASP and ASP.NET technologies which help downloaded malware stay hidden, were also that weren’t properly secured during Web application development proved vulnerable to this type of attack. prevalent. Other widely distributed malware included spyware and keyloggers, both of which send information about a compromised user’s computing and Web surfing habits and personal information—including passwords— back to the malware distributors. The volume of malware successfully propagated via email attachments has declined in recent years. This decline could be related to Web-based malware distribution methods proving so effective, and to the ability of anti- malware products to rapidly detect and block much of the email that contains malware. These factors may have led malware creators to spend more time on malware spread In September 2008, BusinessWeek.com became another via the Web rather than via email. well-known, legitimate website compromised by SQL injection. Hundreds of pages had malicious iFrames redirecting users to a site in Russia where they were unknowingly served malware. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 8 Cisco 2008 Annual Security Report

  • Page 11

    Of the attachment-based malware campaigns of 2008, Mobile Phone Malware: The “Shadow” popular ones included messages claiming to contain delivery forms from UPS or FedEx, or messages claiming Growing Profit Centers Internet Economy their attachment was an invoice, an e-ticket, an e-card (from Hallmark, for example), or video or pictures that were 2008 saw several instances of malware designed for Successful online criminals are making millions or actually executable files. and spread via mobile phones. hundreds of millions of dollars from their enterprises. These profits continue to drive innovation and Another attachment-based malware campaign shut One example is SymbOS/Kiazha.A, a “ransomware” specialization. down the IT systems of three British hospitals, when the Trojan that runs on Symbian OS devices and deletes hospitals’ computer networks became infected with email- incoming and outgoing SMS (text) messages. When Like the legitimate Internet economy they shadow, propagated Mytob malware. it infects a mobile phone, the phone will display the online criminal world has become a global, a message asking the user to send money (to an thriving network of product and service providers and To prosper, malware creators must develop tools that undisclosed location, using a mobile phone recharge consumers doing business together. In the short term, are tough for anti-malware solutions to detect; they are card) to have the device restored to normal function. this specialization and collaboration are making online building more surreptitious malware designed to avoid criminals more nimble and effective. detection by anti-virus and anti-malware programs. One This Trojan is installed on the phone by SymbOS. Multidropper.A, which also installs SymbOS/Beselo, Those launching attacks are often no longer the popular technique is malware that can temporarily go a worm that propagates by sending itself as MMS developers creating the tools. Instead, attackers can dormant. Another is malware code that continually and (multimedia) messages every two minutes to every select from an array of competing and increasingly automatically changes just enough to confuse signature- contact in the mobile phone’s phonebook. It can sophisticated products and solutions. A wide range based anti-malware scanning software. also propagate via Bluetooth, and copy itself to any of well-designed malware offerings is currently for memory card inserted into the phone, allowing it to sale or rent, including: recover from deletion. In another tactic to enhance propagation, SymbOS.Multidropper.A installs • Botnet management and dashboard-type tools Volume of Malware Successfully SymbOS/ComWar.C, which spreads via Bluetooth • Mass blog posting tools Propagated via Email Attachments and replicates and monitors itself to ensure it is not erased from the phone. • Sophisticated volume spamming tools 1000 • Automated webmail account creation tools (including During the last year, malware for mobile phones some that defeat the CAPTCHA feature that webmail was largely circulated in Asia, where the number of 800 hosts such as Yahoo!, Gmail, and MSN use to prevent people who own such devices is significantly higher bots from opening webmail accounts) than those who own personal computers. This makes 600 spreading malware via mobile phones a potentially • Account generators that enable spammers and profitable endeavor for malware creators in that region. scammers to bulk-post to Craigslist 400 • Keylogging programs 200 But in the longer term, the online crime economy may also be on its way to becoming a bureaucracy. The 0 positive side to this: One unavoidable side effect of 2005 2006 2007 2008 becoming more established is a paper trail, which (as of 10/31) may make it easier for law enforcement organizations The volume of malware successfully propagated via email worldwide to track and apprehend more of these attachments declined slightly in 2008 versus 2007. These last two offenders in the future. years represent a 50 percent drop-off relative to the previous two years, in terms of attachment-based attacks. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 9

  • Page 12

    Asprox: Transforming Botnet Trends The Importance of Social Engineering an Old Trojan Botnets are the big “workhorses” that power many of today’s online threats and criminal activities. Botnets Online criminals have developed an array of sophisticated social engineering techniques to entice victims to open consist of thousands of malware-compromised computers. an email or file, or to click on a link or online ad. The use One of the most effective botnets of 2008 was Asprox, and sophistication of social engineering techniques in Those who control the botnets can rent out the processing an old Trojan that was turned into a very sophisticated online attacks continued to grow in 2008, and this trend is power and bandwidth available to these computers, or botnet and used in thousands of SQL injection attacks use it themselves. expected to continue during 2009 with even more—and on legitimate websites. First used several years ago as even better executed—attacks occurring via email, instant a password-stealing Trojan, it was later upgraded to Online criminals are using botnets for pretty much every send phishing spam. Its big transformation occurred in messaging (IM), and mobile devices. aspect of Web-based threats, including spamming, May 2008, when Asprox started updating itself with a One successful technique is creating spam campaigns sending DDoS attacks, infecting legitimate websites, SQL injection tool. hosting malicious websites (such as botsites), and based around “hot topic” news items and current events. This SQL injection tool looks legitimate to users of propagating more malware. Sometimes, these spam emails direct victims to a infected computers, running as “Microsoft Security malicious site that will attempt to download malware to The Storm botnet, enormously widespread in 2007, was Center Extension” (msscntr32.exe). Meanwhile, in the their computers. only a harbinger of what was to come. New, even more background, it is actually using Google to scan the sophisticated, robust, and scalable botnets, such as But more sophisticated campaigns that include extremely Web for Active Server Pages (.asp), which can be Mailer Reactor, Kraken, and an updated, powerful variant clever phishing websites, and where both spam and susceptible to SQL exploits. of Asprox (which had been around in a less able form for websites use social engineering techniques tied to When the SQL injection tool finds vulnerable pages, several years), have also had great success. current events, have also become common. it inserts a malicious iFrame into page content. The These botnets are designed as reusable platforms that can With those campaigns, victims are lured to legitimate- iFrame invisibly redirects a site visitor’s browser to cycle, synchronize, and distribute dynamic attacks. Like looking websites where they are asked to provide personal malsites that try various methods of infecting the victim’s computer with malware and adding it to the many Web 2.0 technologies, they “promote” collaboration information during what appears to be an actual transaction. Asprox botnet. To make it harder to detect to anti- and depend on the network effect. They’re adaptive and However, victims do not receive the goods or services malware programs, Asprox communicates via proxy intelligent, and offer flexibility, redundancy, and security they thought they had purchased. Or, in cases where the server on TCP ports 80 or 82. protocols inspired by modern peer-to-peer (P2P) networks. fraudsters do send something, counterfeit or poor-quality items are delivered—for example, fake pharmaceuticals Cisco data showed that at its peak, Asprox was disguised as brand-name prescription medication. successfully iFrame-injecting 31,000 different websites per day. The more effective email-attachment-based malware distribution campaigns of 2008 also used clever social engineering techniques. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 10 Cisco 2008 Annual Security Report

  • Page 13

    One type involved email claiming to be from UPS or FedEx, New President, New Malware which asked the recipient to review an attached invoice or delivery confirmation to discover what happened to a Current events-oriented email messages convince fictitious package. When the victim opened the attachment, recipients to open and act on the email. In a recent the malware installed itself and let the attacker gain control example, a spam campaign invited recipients to watch a over the infected computer. video of President-elect Barack Obama’s victory speech. Subject line examples included: • Election Results Winner • The New President’s Cabinet? “These virus-laden emails that • Obama Win Sets Stage for Showdown claim they’re from FedEx or UPS The email directed recipients to a fake government- themed botsite. Once there, they were prompted to are really clever. It’s no wonder install an Adobe Flash Player update, which was actually data-stealing malware. Once installed, the malware stole people respond to them!” Link to Active screenshots and passwords, sending that information Malicious URL to a Web server located in Kiev, Ukraine. —Michael Postlethwait, Cisco Security Analyst Another notable campaign occurred around U.S. tax-filing Prompted to install an Adobe Flash Player update time. This one involved email that looked as if it had been sent by the Internal Revenue Service (IRS). It had a very official appearance, and played on the widespread fear of the IRS and worries about not opening or responding to its letters. Only the most savvy recipients realized that the IRS does not send notifications in email, but only uses paper mail sent via the U.S. Postal Service. The continuing popularity of “scareware” can also be explained by viewing it as an example of successful social engineering techniques. Scareware pretends to be anti- malware or anti-spyware scanning software, but is actually malware that is taking advantage of computer users’ fear Government-Themed Botsite The Real America.gov Site of spyware or malware to infect them. The websites these In this example, recipients of a message—which claimed to include downloads are offered from often look extremely credible a link to Barack Obama’s victory speech—were actually directed to a and professional, and often include fake logos and botsite serving up data-stealing malware. endorsements from industry organizations. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 11

  • Page 14

    Beijing Olympics Fake Ticketing Scams One of the most elaborate social engineering Internet scams of 2008 was related to the Beijing Olympics, The site was confusing, and with criminals making a profit of an estimated US$40 to offered minimal help in navigating its offerings. $50 million. People in several countries, from New Zealand to the United States, were taken in by fake ticketing sites You never wondered where you needed to go. Images and text that sold illegitimate or nonexistent tickets to Olympic made everything clear. events. Some individuals paid thousands of dollars for particularly hard-to-come-by tickets, such as those for the opening ceremonies. Information on the Olympics and The biggest offender was Beijingticketing.com, a Olympic events was secondary to Ticket buying was difficult. No professional-looking website that featured the official Ticket buying was easy. You came information about CoSport, the tickets were available on the home company selling the tickets. page. Instead, there was a link to a Beijing Games logo. This fraudulent website was superior to the home page, clicked on the second site that had the tickets. event you wanted and could buy to the official ticketing site, with a better ticketing a ticket. purchasing process and integration with social networking sites like Facebook to virally spread the fake site. Even MSNBC initially believed the site was credible: An MSNBC Forbes Traveler article featured a link to the site. This helped it gain a high search engine ranking, which resulted in ticket seekers who used search engines to look for tickets going to the fake site rather than legitimate sites. You were invited to tell friends about what you’d found through Beijingticketing.com asked users to register—and provide Facebook, StumbleUpon and No attempt was made to take The site was filled with information other linked sites. advantage of social networking. confidential information—before they could purchase and graphics about the Olympics If you found the site, there was no tickets. After registration, users provided credit card and Olympic events. easy way to let others know about it. numbers and “bought” tickets, which they never received. Not only did the scammers net millions of dollars, but they also scooped up thousands of valid credit card numbers for later use or resale to other online criminals. Scam Ticketing Site Official Ticketing Site Fraudulent Olympics ticketing websites, such as beijingticketing.com, took advantage of thousands eager to buy tickets to the 2008 Beijing Summer Olympics. Source: www.beijingticketscam.com. Moriarty Leyendecker Erben LLP. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 12 Cisco 2008 Annual Security Report

  • Page 15

    Spam and Phishing Trends • Email pretending to be from universities with which Spear-Phishing Examples the intended victims are affiliated as current students, Spear-phishing messages currently represent about Cisco estimates that currently, almost 200 billion alumni, or faculty one percent of all phishing campaigns, but are expected messages per day—or approximately 90 percent of all email sent worldwide—can be defined as spam. That’s • Email that attempts to lure the victim into entering login to become more prevalent. This trend bears close double the volume of the previous year, and represents information about their Google Adwords account (not monitoring, because the attacks are becoming more 200 spam emails per day for every Internet user on the only is the victim’s credit card or personal information sophisticated: Criminals are investing time and resources planet. Spam has undergone a significant evolution in the stolen, but often, their Adwords traffic gets redirected in personalizing spam and making the messages seem last year. Massive volumes of pharmaceutical and get- to criminal-run blogs) credible. Why? Because the jackpots are higher when rich-quick spam from botnets remain a resource- and they succeed in obtaining sensitive personal data from • “Whaling” emails, which are extremely personalized processing-intensive issue for many organizations and specially targeted, attractive victims. to target specific top executives service providers. Still, the network protection, anti-spam, and filtering solutions in place at most enterprises have made high-volume, low-sophistication spam more of an annoyance than a security issue. Typical spear-phishing attacks consist of four steps: The spam that does ultimately make it into recipients’ inboxes is becoming ever more dangerous and attractive, By launching malware, hacking into networks or and thus likely to be opened. Newer spam campaigns buying lists from other nefarious online resources, typically include “blended threat” spam messages, which scammers obtain a specialized distribution list of incorporate URLs to entice recipients to click through to valid email addresses. Spear-phishing attacks require criminals to efficiently build appropriate resources malware-distributing or phishing websites. and trick victims into revealing valuable They register a domain and build a fake (but private information. Another type of spam that has become noticeably credible-looking) website to which phishing email more common this year involves targeted phishing, also recipients are directed. known as “spear phishing.” For these attacks, sophisti- They send phishing emails to their distribution list. cated online criminals have been using smaller phishing campaigns aimed at more targeted groups of recipients— Scammers receive login or other account details to great effect. from victims, and steal From: ci@irs.gov data and/or funds. [mailto:ci@irs.gov] Sent: Wednesday, June 06, 2007 1:14 PM To: Earlier phishing campaigns were widespread and high- Subject: Internal Revenue Service Complaint for [case id: #602f41571ba161cc3dc795df7886f000] volume, and typically pretended to be from large banks Mr./Mrs. (IronPort) with a national presence. Then, an increasing number of We regret to inform you that your company is currently being investigated by our CI department for criminal tax fraud phishing campaigns started using the identities of regional due to a complaint that was filled by a Mr. Keith McCall on 05/06/2007 and local banks located near the recipient (and thus Complaint Case Number: MT1CF23A Complaint made by: Mr. Keith McCall involved fewer messages per campaign). Complaint registered against: (IronPort) Date: 05/06/2007 The latest types of spear-phishing campaigns include: You are being investigated for submitting false income tax returns with the Franchise Tax Board. Instructions on how to resolve this issue aswell as a copy of the original complaint can be found on the link • Spam sent via SMS to the mobile phones of recipients in bellow. the same area code Complaint Documents <http://business-complaints.com/Complaint.doc.exe> All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 13

  • Page 16

    Spear-phishing emails often succeed because they mimic messages from an authoritative source, such as a financial Targeted Attacks as a Percentage of Spam institution, a communications company, or some other easily recognizable entity with a reputable brand. 0.45% Unlike more common “mass” phishing emails, however, 0.40% spear-phishing attacks rely on specific (usually stolen) 0.35% information to craft a more personalized message— 0.30% one the recipient is more likely to open and respond to. Percent The personalized approach of spear phishing, combined 0.25% with email reputation hijacking, in which criminals use 0.20% a legitimate email provider’s infrastructure to send 0.15% messages, makes it more difficult to weed out these 0.10% emails via standard anti-phishing technologies. 0.05% In many cases, online criminals rent or steal lists of 0.00% valid email addresses, and can therefore personalize 12/07 1/08 2/08 3/08 4/08 5/08 6/08 7/08 8/08 9/08 10/08 outgoing messages. Consequently, even savvy Internet Month users—conditioned to ignoring the less-sophisticated phishing messages sprayed to millions of people at the same time—can be lured into handing over login names, passwords, and other sensitive information. For example, online criminals have been sending spear- New Spear-Phishing Attacks phishing messages that appear to be from entities such as: • The Internal Revenue Service, explaining that the Value per Threat Attempt recipient or the company is being audited. • The Better Business Bureau, which has received a “complaint” about the recipient’s company. Traditional Spam Campaigns • U.S. district courts or tax courts, notifying recipients that they are being subpoenaed. These messages look authentic and typically ask recipients to rapidly respond to the inquiry, which usually includes an attached “explanatory document.” However, Number of Threat Emails in Each Campaign when opened, this file actually launches malware in the background to take control of the recipient’s computer or network, or to install a keylogging program. Spear-phishing campaigns are sent to fewer recipients, but are more likely to offer higher returns to criminals when recipients do respond to them. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 14 Cisco 2008 Annual Security Report

  • Page 17

    Email Reputation Hijacking In email reputation hijacking, real email accounts with The appeal of reputation hijacking has led to growth in for just under one percent of all spam worldwide, but major legitimate webmail providers are used to send the number of commercial tools available to spammers. constituted 7.6 percent of all these providers’ mail. out spam. Taking advantage of the webmail provider’s These tools are aimed at making it simpler for The average spam rate from each webmail provider positive reputation offers increased deliverability: It spammers to create accounts, defeat CAPTCHAs, post, rose significantly for a period of time after tools to take makes the spam harder to detect and block, since it and rotate IP addresses to target webmail providers advantage of their systems became available. has the webmail provider’s headers and formatting, and like Gmail, Yahoo!, and Hotmail, as well as sites like For example, in January 2008, Russian hacker “John anti-spam solutions cannot block the mail servers of MySpace, Craigslist, and blogs. Wane” defeated Yahoo!’s CAPTCHA. This led to an large webmail providers like Yahoo!, Gmail, and Hotmail. Cisco estimates that during 2008, spam due to email HTTP spike, followed by a three-month SMTP blitz. In reputation hijacking from the top three webmail May 2008, Google’s CAPTCHA was broken, which led to providers—Microsoft, Gmail, and Yahoo!—accounted an enormous spike in account creations. In August 2008, “John Wane” released AOL CAPTCHA-breaking code. Email Reputation Hijacking Tools Email reputation hijacking tools for the major webmail providers are commercially available and easily obtained. These tools were used frequently in 2008 and spam originating from these webmail providers increased significantly. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 15

  • Page 18

    Data Loss Despite best efforts, security incidents related to data loss are on the rise. Data loss can occur because of the physical loss or theft of systems and storage devices or the accidental sharing of information in an insecure fashion. Online criminals are using malware to steal consumer and company data online. Hackers are also getting their hands on data by breaking into insecure or weak systems and devices. And sometimes, insiders are the culprits. 16 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 19

    More and more businesses are recognizing that their an employee’s car. The laptop was encrypted, but the Data Loss Issues on the Regulatory Radar data is a precious asset that must be protected. flash drive was not; the latter contained potentially Data breach notification legislation now requires that PricewaterhouseCoopers’ 2008 Global State of sensitive business information, according to a company companies report when sensitive data is potentially Information Security Study reports that many more spokesperson. lost—such as when a laptop is stolen. organizations are encrypting “sensitive information not • In the summer of 2008, Verified Identity Pass (VIP), a just in laptops, but also in databases, file shares, backup Currently, 42 U.S. states and the District of Columbia have vendor of the U.S. Transportation Security Administration tapes, and removable media.” And many have made data breach notification laws on the books, or legislation that operates a Registered Traveler program under the “significant strides in advancing Web/Internet capabilities,” pending approval. While state laws vary, they generally brand Clear, temporarily misplaced a laptop that had including content filters, website certification/accreditation, follow the California Security Breach Information Act been reportedly locked in an office at San Francisco and secure browsers. The report also cites increased (SB-1386), which requires organizations that electronically International Airport. The laptop contained unencrypted use of technologies that help protect wireless devices, store personal data about customers to inform those personal information for 33,000 customers, but according and tools that can discover unauthorized devices or individuals if the company knows the security of that to the vendor, none of the data was compromised. prevent intrusions. information has been compromised. • The U.S. Veterans Affairs (VA) Department, which suffered According to the Privacy Rights Clearinghouse’s Meanwhile, the U.S. government is attempting to address severe embarrassment in 2006 when a laptop containing Chronology of Data Breaches, since January 2005, more these notifications at the federal level and consolidate millions of veterans’ records was stolen, had its new than 230 million records have been compromised due to the variances in state laws—perhaps even strengthening security policies put to the test in spring 2008, when security breaches. Cisco research shows that inadequate some laws. It is also trying to align federal legislation with another of its laptops was stolen from an employee’s Texas data security can have significant consequences for approved or pending legislation in other countries. apartment. This time, the data was encrypted, no one organizations, including business disruption, reduced without proper authentication could access the computer, In addition, there are increasing laws and regulations to productivity, and increased operational expenses—and and the VA knew which piece of equipment was missing. handle the sharing of sensitive data. New U.S. compliance those are on top of the obvious loss of sensitive data. The employee also did his part by immediately reporting regulations and market-driven industry best practices Data loss related to the loss or theft of equipment is an the theft to the VA and local authorities. were released in 2008 that attempt to focus on stronger enormous problem for businesses and individual users. protection and increased enforcement for data loss • The U.K.’s Ministry of Defense reported a serious breach Ponemon Institute recently reported that the number of violations. The federal government released the Red of security when a laptop containing unencrypted data lost laptops at some medium-sized and large airports has Flags provision of the new Fair and Accurate Credit related to 600,000 prospective military recruits, including been reaching more than 600,000 annually. More than half Transactions Act, which spans multiple industries and some financial and passport information and medical of the laptops are never reclaimed, as many people hold requires businesses that provide services before billing to details, was stolen from a military recruitment officer’s car. no hope their laptop will be found. Therefore, they often implement an identity theft prevention program. The initial do not bother taking any steps to attempt to locate and Fortunately, very few incidents of equipment loss or theft enforcement period was November 2008, but it has been retrieve them. result in information being passed on to criminals with the delayed to May 1, 2009. expertise to profit from accessing and using the comp- The following incidents that occurred in 2008 are related In October 2008, the state of California passed two privacy romised data. These are usually simple thefts, with the end just to laptops: laws, SB-541 and AB-211, that attempt to augment existing goal of quickly reselling the equipment, which is wiped medical privacy compliance regulations by focusing on • A laptop computer containing personal information— clean of data to conceal the fact it has been stolen. the enforcement of unauthorized access to patient health including names, addresses, and employee identi- information, negligent disclosure of patient records, and fication numbers—for approximately 13,000 workers illegal use of medical information for financial gain. The of a global pharmaceutical company was stolen from Health Information Trust Alliance will also release the HITRUST Common Security Framework (a market-driven All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 17

  • Page 20

    best practice) in January 2009. This framework is built environment. In fact, multiple security incidents in 2008 Recycling Risks upon industry standards such as ISO 2700x, B/S 7799, involved organizations that were considered “compliant,” Recycling of electronics equipment is becoming more PCI, and the NIST 800 series. but were compromised by exploits not covered by common. In fact, some countries already charge buyers Individual U.S. states are also imposing their own laws to compliance requirements. Many compliance procedures a recycling fee when they purchase any electronics item. mandate encryption of personal information sent over do not and cannot address today’s array of applications, All or part of that fee may be reimbursed if, at the end of the Internet by businesses located in the state. New laws technologies, tools, and the related security vulnerabilities its useful life, the item is taken or sent to a recycling center. and regulations require, among other things, encryption that are increasingly being targeted by threats. Instead, Organizations—looking to recoup those fees, comply with of personal information on laptops, PDAs, and portable compliance measures are intended to help organizations laws and regulations, or simply be more environmentally media (including flash drives); encryption of personal achieve specific objectives that mitigate only certain responsible—are likely to recycle more often. Yet while information transmitted over the Internet; development security risks. more organizations are recycling their “e-waste,” many and publication of Social Security number (SSN) privacy The market-driven Payment Card Industry Data Security aren’t taking sufficient precautions to make sure those protection policies; and specific measures to protect the Standard (PCI DSS), for example, focuses on the protection items have been wiped clean of sensitive data. confidentiality and security of employee SSNs. of cardholder data during processing, transmission, or Once equipment destined for recycling is sent away, there While such legislation obviously benefits those whose data storage. It is a detailed standard compared to other is no telling where it may go or what will happen to any data has been compromised, disclosing a security breach may compliance regulations. However, it does not entirely that can be extracted from it. Press reports indicate that leave an organization subject to negative media coverage mandate a strong level of security, as it must balance its some devices have ended up as far afield as Indonesia and possibly cause long-term reputation damage. This can strict requirements with a risk-based approach that can and West Africa, where salvaged data is sold at bargain- lead to a drop in the confidence of users and customers, apply to both small and large organizations. The industry basement prices. who may be inclined to take their business elsewhere. recently updated PCI DSS to version 1.2. In addition to merging both versions of the requirements and testing Many organizations do not make the IT department Many laws and regulations carry significant statutory validation steps, the revised standard provides new responsible for (or at least involved in) the electronics penalties for violations as well as the possibility of deadlines around Wired Equivalent Privacy (WEP) equipment recycling process, instead leaving it to other businesses facing private rights of action for noncomp- replacement and adopts a risk-based analysis approach departments, such as facilities management. However, liance. As a result, an increasing number of businesses are so to help smaller businesses comply. it is unlikely that personnel outside of IT will be aware of using encryption and other access control technologies to the importance of degaussing hard drives and otherwise help ensure compliance. To address the array of compliance best practices safeguarding potentially sensitive data on defunct devices and regulations, organizations will likely plan larger before those devices end up in a recycling facility. The Limitations of Compliance IT governance, risk management, and compliance Consequently, organizations without clear, security-oriented (GRC) programs. Although achieving compliance is While many regulatory standards attempt to help protect policies for how to and who should handle this process important, organizations must remember that many best user data, compliance cannot be a security placebo. within the organization may put sensitive data at risk. practices and regulations apply only to protecting certain Many companies have made great strides to achieve information—for example, company financial information Identity Theft compliance measures, but the sense of urgency that often (Sarbanes-Oxley), patient medical information (HIPAA), and surrounds compliance demands should not become Identity theft continued to rise during 2008 and shows no personally identifiable information (Basel II, and U.K. and a distraction from other, crucial threats to security. By signs of slowing down. Many online criminals have been EU data laws). Therefore, in addition to regular compliance focusing almost exclusively on compliance and aligning successful at using social engineering tactics that feed reviews, organizations will often conduct top-down gap procedures to meet those requirements, organizations on the trust of others, and allow sensitive personal data to analyses to improve existing procedures and proactively can lose sight of the rapidly evolving risk and threat be harvested, ranging from Social Security and driver’s meet current and emerging threats. license numbers to complete medical histories. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 18 Cisco 2008 Annual Security Report

  • Page 21

    Collating data from a variety of publicly available sources— terms of reputation damage and financial loss, particularly look to prevent data loss, reduce the potential for identity including user profiles on popular social and professional in instances when the trust of many consumers has been theft, limit opportunities for insiders to engage in criminal networking sites—makes it easy to pull together enough compromised. For example, sophisticated phishing activities, and comply with regulatory standards. They information about a person’s identity to create a scheme techniques can dupe users into believing they are also require technology that supports collaboration that either takes advantage of the individual, or of people interacting with a trusted source—such as an individual, among their remote workforce using Web 2.0 tools and who they know and who trust them. charity organization, bank, or online retailer—via spam applications, as well as with other organizations. For victims of identity theft, the risks have increased emails and with legitimate-looking but fake websites. Today’s identity management solutions have advanced significantly: Just one security breach—whether it is Other data-gathering opportunities for today’s identity well beyond easy-to-compromise usernames and keylogger malware invisibly downloaded to their home thieves include: passwords required to access networks or applications. computer by a compromised website, or hackers cracking • Social networking—A rich trove of personal Secure, personalized user profiles may be created. And into the customer database of their favorite retailer information, including phone numbers, addresses, full before access is approved, users may have to be verified thousands of miles away—can compromise their names, and birthdates, is available on user profiles through one or more methods, including tokens or smart personal information. posted on social networking sites such as Facebook cards. Some organizations use biometrics, including And now, stolen personal information is being bundled and MySpace, and additional personal information such fingerprints and iris scans, to authenticate users. and sold to criminal elements around the world—such as mothers’ maiden names on sites like Ancestry.com. Transparency is another theme in modern identity as organized crime rings or even hostile governments. • File sharing and peer-to-peer software—When management. Organizations want technology that allows Those who tap this market use the information not only users allow friends and associates to access certain them to monitor user activities from sign-on to sign-off. for profit, but also to sell to those who use the information files, such as MP3s, other files on their computers can be They need solutions designed to set boundaries on the for terrorism-related activities, such as creating fake easily compromised. “Access creep” is also a growing amount and types of information and other resources users passports and other travel documents, or laundering problem in collaborative work environments, with people are permitted to access. In addition to making it easier to money for terrorist cells. being allowed to view too much information, including track user activity, identity management technologies now company secrets and information about coworkers. on the market allow organizations to set consistent user Targeting the Masses policies and conduct auditing and reporting to help assure Although the instances of highly targeted phishing • RFID tags—Some concealable readers can read compliance with regulatory standards. campaigns are growing in number, most online criminals radio frequency identification (RFID) tags from a short distance (up to a few feet) to gather data from credit and There is strong demand for identity management solutions looking to commit identity theft are not going out of their that are complete and highly effective but also easy to way to target specific individuals or groups—at least, other types of cards. RFID tags can be cloned, and the equipment required to do this is available. This has clear use. That’s why more organizations are looking to “single not yet. They are simply trying to snare as many people sign-on” solutions that simplify the process of verifying as possible. security threat implications: RFID technology is used in everything from building access cards to passports. user identity and give users access to the information and The Internet provides ample opportunities for identity applications they need. But while identity management thieves looking to target the masses. According to a report Rethinking Identity Management technology has advanced dramatically in recent years, released by the FTC in February 2008, some 64 percent the industry continues working on developing solutions Given the concerns around data loss and identity theft, the of fraud complaints in 2007 related to incidents where that provide even greater security and user monitoring recent resurgence of interest in identity management is the method of initial contact was an Internet solicitation, capabilities—and don’t hinder workforce productivity. not surprising. Many leading companies are overhauling such as email. the security platforms they have long relied on and The economic impact to individuals who are victims of adopting new identity management technology. They identity theft is obvious. But businesses also suffer in All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 19

  • Page 22

    The Human Factor People are still the weakest link in the security chain. But their capacity to learn and modify their behavior in response to information means they also represent an area with a great opportunity for improvement. Attacks on websites and corporate networks continue to increase in sophistication, and online criminals are growing proficient at duping even the savviest or most cautious of users. 20 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 23

    Human Nature Invites Risk Technology solutions (such as anti-spam tools and Remote Working, Social Networking: outgoing email monitoring) can be helpful for proactively Plain and simple human error—such as a CEO opening an Opportunities and Risks mitigating some risks and preventing widespread damage email that appears to be from a trusted source, but is really Having a mobile workforce can significantly improve from certain types of attacks. Providing ongoing threat a well-disguised “whaling” attack, or poor judgment, such business productivity, and keep workers happy by allowing education and training for employees—and building their as engaging in e-commerce on a website that does not for better life-work balance. A global workforce can be awareness about security risks and the importance of have a valid security certificate and is a front for a scam— very cost-effective, allowing for localized service to safeguarding data—remain important security defense is often what triggers the release of malicious attacks or customers in different regions and faster entry into new measures for organizations. But they have their limitations. leads to identity theft and other fraud. markets. And many workers today (especially young, highly Technology can create a false sense of safety, and neither Human carelessness—for example, losing an employer- wired “Generation Y” workers) love using new tools and that or education can address a broader security problem: issued laptop or inadvertently posting a company’s technologies that make work and life easier and more fun. human nature. sensitive information on a blog—can also quickly turn Organizations are equipping their remote workforces with By nature, most people are curious, eager to communicate, into a reputation-damaging event, and cause significant the collaborative tools and mobile devices they need to and interested in good deals or attractive “freebies.” Quite financial loss for an organization. Email address errors are do business anywhere, anytime. But by doing so, they often, they also are overconfident that they are not the another common human mistake-based security problem also create security risks. For example, in its Emerging type of people who would fall prey to trickery or scams; that can easily result in highly sensitive information being Cyber Threats Report for 2009, Georgia Tech Information this aspect of human behavior is a key element in making sent to the wrong people. Security Center warns that as Internet telephony and online crime work. Even the security-conscious U.S. military is not immune mobile computing—which are essential to remote Online criminals thrive by taking advantage of Internet workers—handle more and more data, they will become to such blunders: In 2008, it was reported that United users’ trust and human nature. Countless people continue more frequent targets of online crime. The report predicts States Air Force (USAF) personnel inadvertently sent email to be lured to malware-distributing or phishing websites that criminals will be “drawn to the VoIP medium to engage messages intended for USAF personnel stationed at Royal by emails containing URLs. Given that more than 80 in voice fraud, data theft, and other scams—similar to the Air Force Base RAF Mildenhall in Suffolk, England, to a percent of spam messages now contain URLs, it is not problems email has experienced.” tourist website with a similar email address. The maintainer difficult to see why this hit-or-miss approach succeeds. of that site, which is intended to promote tourism in Meanwhile, the line between technology use for personal And with potent and self-propagating malware available, Mildenhall, notified the USAF several times about the and professional purposes is becoming increasingly even a relatively small number of infected users can infect emails but was told not to be concerned. Only when blurred. Recent Cisco research revealed that 44 percent many more. officials were notified that flight plans for a presidential visit of employees share work devices with others without were received did they become—reactively—alarmed. supervision, and 46 percent said they transferred files between work and personal computers when working at home. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 21

  • Page 24

    Cisco’s recent research into security perceptions and Inadequate or insufficiently communicated appropriate online behavior of remote workers in several countries use policies will have to change. With popular sites (including Brazil, France, India, and the United States) such as MySpace and Facebook likely to remain highly showed that using work computers and devices for vulnerable to hard-to-detect malware such as Koobface, personal use is a widely accepted practice today. A organizations will need to pay more attention to on-the-job primary reason for this casual attitude cited by survey social networking by employees. Koobface works to turns respondents: The belief that their employer does not infected users’ computers into botnet nodes. The worm mind. What’s more, many users download certain tools searches for cookies associated with a social networking and applications in an effort to be more efficient in their site and, once located, modifies them and embeds jobs, but can wind up derailing overall productivity by malicious links on a user’s profile. Others viewing the unknowingly creating a convenient inroad for a threat. profile assume the links were put there by the user. Trusting the source, they click on the links and also become A report by Telework Exchange Elements of human nature, such as being curious and having the desire to connect with others, also create infected with malware. (Instructing workers to clear their and Sprint Nextel on wireless risk for organizations, which can expect more of their cookies on a daily basis can help combat this problem.) employees (mobile or otherwise) to engage in social and Meanwhile, the laptops, mobile phones, PDAs, and data Internet usage among U.S. professional networking online while on the job. Some storage devices that both remote and onsite workers are government employees companies are encouraging this activity—using such using to engage in business and personal activities are outlets for their own marketing, PR, and HR initiatives. providing myriad points of entry for spam, viruses, and revealed that 33 percent of Even the microblogging utility Twitter is now being used malware. They also offer endless opportunities for loss of by leading companies to share product news and offer intellectual property and other data. teleworkers, and 11 percent of special deals to those who sign up for the service. Workers may be insufficiently aware of the unprotected IT workers, were not familiar with Providing proactive and thorough user education, and nature of mobile phones, the need to use privacy screens setting clear policies about social networking and other while doing sensitive work in public places, and taking security guidelines for using online activities while at work, is good practice. However, extra care with easily misplaced mobile devices. A thumb wireless Internet. many organizations either fail to set appropriate use drive can now contain as much as 64 GB of data, and the policies or do not communicate them to users or internal U.K. Ministry of Defense recently had to admit to several resources expected to help enforce such practices. instances of sensitive information lost by officials moving A report by Telework Exchange and Sprint Nextel data physically between locations, including a classified on wireless Internet usage among U.S. government report on Al Qaeda, which was left on a train. employees revealed that 33 percent of teleworkers, and 11 percent of IT workers, were not familiar with security guidelines for using wireless Internet. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 22 Cisco 2008 Annual Security Report

  • Page 25

    Online scammers have been running so-called “419” Using Social Networking (named after the section of the Nigerian penal code, where such scams often originate) or advance-fee fraud— and Web 2.0 Sites for a confidence trick that has been around for decades— on websites such as LinkedIn and Craigslist. A typical Online and Offline Crime Craigslist scheme: The criminals use legitimate but outdated house-for-rent postings from other Craigslist Social networking websites such as Facebook, MySpace, users to dupe potential renters into wiring money to the Bebo, LinkedIn, Orkut (extremely popular in Brazil), and “landlord,” who claims he or she had to suddenly move vKontakte (in Russia) have all been linked to spam and overseas, and must rent out his or her house immediately— malware attacks. The Web 2.0 widgets and different types and cheaply. of potentially vulnerable content that users can add to their pages may make it especially easy for malware creators In some cases where an address for the home is provided to exploit these websites in the year ahead. in the ad, victims have been told by the scammer that while a property tour was not available, they should drive Yet not all criminal uses of social networking and Web by the house to have a look. While there may be plenty of 2.0 sites take place exclusively online. Canadian and U.S. red flags visible in retrospect, for many people seeing is anti-fraud organizations reported a disturbing trend that believing, which is why this scam has been successful. surfaced in 2008: A sharp uptick in phone calls used in extortion schemes targeting senior citizens, which have In another recent social engineering scheme involving resulted in victims paying thousands of dollars to criminals. Craigslist, a resourceful bank robber enlisted “help” for a The caller relays a fictional story about the senior’s teenage robbery getaway through false advertising. The suspect or college-age grandchild needing to make bail in Canada— posted a listing on the site advertising a road maintenance likely gathering personal information to use in the scam, project that would pay US$28.50 per hour. Around a dozen such as the names of victims’ grandchildren, from profiles unsuspecting decoys-to-be who applied were asked to posted on social networking and other Web 2.0 sites. show up for work at a local bank wearing specific attire— a yellow vest, safety goggles, a respirator mask and, if It also works the other way around, with online criminals possible, a blue shirt. “borrowing” elements from the physical world to help make their online scams appear credible. They did as instructed, but when they arrived at the jobsite they found no work to be done and no contractor. At the same time, the robber (wearing the same gear as the decoys) wrestled a bag of cash from an armored truck guard and made a clean getaway—leaving the police to sort through a dozen look-alike suspects. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 23

  • Page 26

    Insider Threats External threats such as Web-based malware and hacker intrusions may be more numerous, but organizations should never ignore the significant security risks posed by insiders. Insider threats can be even more damaging to a company’s reputation and financial well-being than ones that originate outside the organization. 24 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 27

    Adding to concerns about insider threats, larger entities installed on point-of-sale servers at the targeted Financial Crisis May such as competing corporations and hostile governments restaurants. Over the course of seven months in 2007, Heighten Insider Risk (as well as organized crime) have been placing agents data was collected from thousands of credit and debit Due to the global financial downturn, Gartner analysts are within organizations they wish to compromise. The extreme cards. At just one location, more than 5000 cards were predicting large IT budget cuts as well as hiring freezes volatility and deep uncertainty felt throughout the global compromised, resulting in at least US$600,000 in losses and layoffs. If workforces are to be cut in response to the economy in the latter half of 2008—which will likely carry for the financial institutions which issued the cards. financial pinch being felt by many organizations, many well into 2009—is another reason why insider threats • There was even more bad news for the U.S. mortgage employees could become “disgruntled about-to-be- can be expected to remain a major security concern for industry in 2008 when an investigation of mortgage ex-employees.” And disgruntled employees could offer businesses of all types. brokers in the state of Florida, conducted by The Miami many opportunities for online criminals to gain access to This year saw a rise worldwide in the number of instances Herald, revealed that thousands of licensed brokers sensitive data, passwords, and the IT infrastructure. of fraud, hacking, and identity theft by insiders—or those had criminal records that should have been discovered For instance, consider the World Bank. According to some who were able to compromise physical security controls during mandatory background checks. Regulators must reports, an IT consultant infected the computers of several to gain access to an organization’s networks. Underscoring approve licenses for mortgage brokers in Florida, where coworkers with keylogging software, gaining the ability this trend are the following incidents reported in the background checks have been required since 2006. to compromise several of its servers. As of mid-October past year that were not discovered until after significant According to the newspaper’s report, more than 10,500 2008, the World Bank denies that sensitive information was financial damage had occurred: people with criminal records were approved to work in compromised, but this story showcases the vulnerability • In January 2008, a trader for French bank Société the mortgage profession. of institutions and organizations that were otherwise Générale admitted that by engaging in unauthorized The investigation uncovered an estimated US$85 perceived as robust and trustworthy. stock market deals, he had caused a €4.9 billion loss for million in losses due to fraud, identity theft, and theft With many companies globalizing their workforces, we are the financial institution. The employee used knowledge of savings and homes involving licensed brokers. In increasingly living in a single, integrated world economy. and experience he’d gained through previous work in addition, several brokers who committed fraud were— Employees of American banks and IT firms may be working the bank’s risk management office to conceal his losses remarkably—allowed by regulators to keep their license. out of call centers in Asia or Europe. Making sure the through falsified transactions. With public records and the current technologies security policies put in place are usable in the context The trader’s fraudulent activity—which required a available to anyone with an Internet connection, of local culture, but also work within the global security breach of five layers of control and included the theft of conducting background checks is a relatively simple, policies of a multinational organization, is crucial. In addition, computer access codes—was discovered by auditors inexpensive, and quick control that can identify potential as companies cut costs they may increase their depen- looking into an error made by the bank’s chairman and security issues. dence on teleworkers and consultants. This can be CEO. It is believed this trader acted alone, and was • In October 2008, an onsite IT contractor for Shell cost-effective, but requires additional security policies motivated not only by the desire for personal gain, but to Oil was caught stealing information about current and implementations to work securely at the edges of an enhance his trading reputation within the organization. and former U.S.-based employees from a company organization’s network. • In May 2008, three men (with direct access to the database. According to Shell, the contractor used Social hardware targeted in the crime) were charged with Security numbers belonging to four employees to file hacking into 11 cash register terminals and stealing fraudulent unemployment claims. After discovering credit and debit card numbers from customers at a the breach, Shell had the contractor removed from the popular U.S. restaurant chain. A “packet sniffer”—a premises. The company dropped its contract with the computer code designed to capture communication associated vendor and alerted its employees of the between computer systems on a single network—was breach. The Texas Workforce Commission and local law enforcement are investigating the incident. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 25

  • Page 28

    Issues of Trust Many users believe they enjoy the same levels of data security they enjoyed in the past, when transactions and related data existed primarily in the physical realm. They trust that organizations they willingly give their personal information to will do everything possible to safeguard it. And they believe the equipment and services of providers they know and trust are secure. 26 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 29

    Users can put their information at risk of exposure in more Another ongoing risk for organizations is people. For ways than ever before, whether by tapping into a local instance, employees can lose equipment, which can coffee shop’s unsecured Wi-Fi network (easily sniffed compromise sensitive data. Insiders looking to commit by identity thieves) or making a purchase from a national fraud also have more incentive today: The increase in data retailer that relies on archaic data storage methods (easily density makes attacks originating from within an organization compromised by hackers). much more profitable. Properly placed devices designed to Organizations can be just as naïve about their own security. sniff or collect sensitive data, or the copying of data that an They may put too much trust in existing protocols, services, employee has legitimate access to, are known to be at the and components, or may not do enough to validate and root of several of 2008’s high-profile security incidents. monitor trusted relationships through available methods Even hardware can pose a threat: Counterfeit chips such as certificates, monitoring, and testing. Meanwhile, inserted into computer equipment can obviously put some businesses compliant with certain regulatory or sensitive data of individual users, businesses, and industry standards assume meeting these standards governments at risk. It may sound like fodder for a spy ensures adequate security. novel, but there have been reports of criminals (and even Ignoring known weaknesses Ignoring known weaknesses is another problem for foreign governments) finding opportunities along the organizations. Consider that much of what was “new” in global supply chain to embed counterfeit components is another problem for the way of threats during 2008 essentially came down into devices. organizations. to online criminals exploiting old problems—unpatched While some experts say the security threat is overblown, systems, weak security policies, and known vulnerabilities counterfeit components can provide the “back door” By not addressing and in the core infrastructure of the Web. By not addressing that external parties need to access a user’s personal and patching existing issues, organizations cannot information or monitor their communication. They are also patching existing issues, adequately prepare to combat new threats. extremely difficult to detect and can be costly to address. organizations cannot The threat against network operating systems has grown While software can be patched, counterfeit components substantially over the past decade and recently, there must be removed one machine at a time. adequately prepare to has been a notable surge in criminal activity related to the exploitation of fragile networks. To fortify the networking combat new threats. and IT systems that make up their critical infrastructure, many organizations are now making such upgrades a part of their security strategy. In fact, some are viewing annual upgrades as being only a bare-minimum effort, and are conducting upgrades twice a year or more. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 27

  • Page 30

    New Tactics Erode Trust Online criminals look for any and all favorable tactics to take advantage of users’ trust—hence, the growing popularity of various kinds of reputation hijacking. In 2008, many leading companies with well-known and Percentage of All Webmail Spam Broken Down By Major Provider trusted brands had their reputations compromised Percentage of All Webmail Spam Broken Down By Major Provider by these attacks. Criminals successfully hijacked 100% reputations by: • Creating highly credible spam that appeared to come from a real company—both visually and by spoofing 80% header information. Recipients were directed to Percentage of all Webmail Spam legitimate-looking websites that were clever fakes. Other • Overcoming security measures designed to avoid 60% AOL the mass creation of webmail accounts from top Hotmail webmail providers with trusted reputations. Once Gmail criminals gained the ability to create large quantities Yahoo! 40% of webmail accounts, they used them to send out massive amounts of spam, which was more likely to get through anti-spam filtering systems due to the legitimate webmail sender address. 20% • Poisoning DNS caches from local Internet providers so that typing in the legitimate URL would lead to a malicious site where users would provide sensitive 0% 7/25/07 8/08/07 8/22/07 9/05/07 9/19/07 10/03/07 10/17/07 10/31/07 11/14/07 11/28/07 12/12/07 12/26/07 1/09/08 1/23/08 2/06/08 2/20/08 3/05/08 3/19/08 4/02/08 4/16/08 4/30/08 5/14/08 5/28/08 6/11/08 6/25/08 7/09/08 7/23/08 8/06/08 8/20/08 9/03/08 personal and financial information. • Inserting malware-downloading iFrames into thousands of legitimate websites (including those Date of major retailers and news organizations) through SQL injection and cross-site scripting, among other methods. A recent Cisco study estimates that 20 The average spam rate from each webmail provider rose significantly for a period of time after tools to take advantage of their systems became commercially available. percent of all legitimate sites have been tainted by this type of attack. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 28 Cisco 2008 Annual Security Report

  • Page 31

    Privacy and Trust Violations Universities, for example, gather and maintain a large amount of detailed personal and financial information More consumers are learning that their privacy may not related to their students and alumni. Last year, it came to be well protected by sources they trust. Businesses and light that many universities have been sharing this infor- organizations are freely sharing consumer information mation with outside companies, including banks and credit with third parties for advertising and marketing purposes. card providers—a practice often in direct opposition to Often, they do not disclose that fact to consumers (or at privacy protection policies, statements, and information least not as clearly as they should). Even when they do provided directly to students. spell out policies, users may not read them (thoroughly or at all) before clicking “accept.” Florida State University recently came under fire for providing names and addresses of students and alumni Sometimes, consumer data is put at risk when everyone to Bank of America for a credit card promotion. The in an organization does not fully understand the privacy Consumer Warning Network (which obtained a copy of the protection practices, or when third-party vendors trusted contract) uncovered that, as part of the deal, the university with sensitive information are not aware of or do not follow receives a portion of every dollar charged by students security policies. and alumni on the credit cards, which feature the school’s IT and privacy departments that establish strict policies colors and logo. Florida State University reportedly is around the use of customer data may find other depart- guaranteed to receive US$10 million over several years, ments undermining those directives. A 2008 Ponemon and the money is being paid directly to the Seminole Institute study of executives shows that security and Boosters, a private entity that raises funds to support the privacy officers responsible for protecting consumer data school’s athletic program—including paying coaches’ gathered by their organizations are clearly at odds with salaries. their own marketing departments, which share the same The irony of this situation: While doing this deal with Bank data (including email addresses) with external parties. of America, Florida State University was simultaneously engaged in a media campaign warning its students of the dangers of credit card debt. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 29

  • Page 32

    Vulnerabilities Vulnerabilities exist in many technologies. Criminals take advantage of these weaknesses to install malware on computers and devices, gain control of computers and networks, and profit by making them parts of botnets or stealing sensitive data stored on them. 30 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 33

    To lower the risks of having criminals gain control over their One result is that, while the overall number of disclosed Web Vulnerabilities systems, IT professionals and individual users work to find vulnerabilities is rising, the number of “zero-day” With the Web being used by more people for more patches, fixes, and upgrades for the products and systems vulnerabilities (vulnerabilities for which there is no patch purposes in more new, untested ways, vulnerabilities along they use. It can look like a race between criminals looking available when exploit code is made public or discovered the entire Web ecosystem—including browsers, Web for new or more attractive vulnerabilities to exploit and in the wild) in products such as major operating systems applications running in those browsers, servers, and some users trying to keep their systems patched and as secure seems to be declining. of the underlying infrastructure of the Web—continue to as possible. Another vulnerability trend is that many attacks now grow in number and importance. However, it can sometimes feel onerous to users to find and use a combination of multiple exploits that each target And it’s not just that new ways of use are creating new install patches or deal with the hassle of upgrading, and different weaknesses to increase the attack’s access vulnerabilities. Many known vulnerabilities in Web-based they may fall behind in keeping their systems and products and control of the system. These combinations used in tools and technologies continue to be exploited by online patched and upgraded. This can be a real boon for cross-vulnerability attacks can vary widely, depending on criminals. Some high-profile Web-based technologies criminals, as certain longstanding but not-always-patched what operating system and programs are running on the known to have vulnerabilities include: vulnerabilities can offer easy ways of infiltrating systems. targeted system. Adobe Flash Player. When users click on and view a The types of vulnerabilities most often exploited have malicious Flash file on a website or in an email, this can changed over the years. Certain vulnerabilities are now trigger the execution of arbitrary code with the privileges more likely to be patched (sometimes automatically) as vendors have developed systems to both disclose and release patches for them. Cumulative Annual Alert Totals In fact, Cisco found that the number of reported 7000 vulnerabilities in 2008 increased compared to 2007, 6500 growing by 11.5 percent. This continues the trend of previous years, and shows that vendors are more actively 6000 reviewing, identifying, and correcting vulnerabilities in 5500 their products. They’re also more often collaborating with 5000 2005 2006 security researchers to do so. 4500 2007 4000 2008 According to the July 2008 IBM Internet Security Systems Alerts X-Force Trend Statistics report, security research 3500 3000 organizations are finding nearly 80 percent of critical vulnerabilities. This correlates with Cisco information, 2500 which indicates that around 80 percent of critical 2000 vulnerability disclosures are coordinated with vendors of 1500 the affected products so that they can release patches or 1000 updates at the time of disclosure. 500 0 1 2 3 4 5 6 7 8 9 10 11 12 Month The number of reported vulnerabilities in 2008 increased compared to 2007, growing by 11.5 percent. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 31

  • Page 34

    of the user. If the user is logged on to their computer with Media players. Many popular media players used to an admin account, the attacker could execute code that play multimedia content that is either downloaded from completely compromises the system. Widespread attacks the Internet or embedded in webpages have proven using this vulnerability were conducted in April and May to be vulnerable to exploits. Vulnerable players include 2008. Adobe released updated Flash Player software in RealPlayer, Windows Media Player, Adobe Flash Player, response, and multiple vendors updated their security and QuickTime. For online criminals, media players can settings and tools to stop this exploit. be especially attractive to try to compromise, since WordPress. In March 2007, an entire version of this widely users are conditioned to receiving messages that they used blog-creation software was compromised when should upgrade their media player to be able to play online criminals gained user-level access to a server different kinds of content or for security. Sometimes these “With the increased complexity hosting the latest official release of the software. They messages are legitimate, and may be ignored due to inserted malicious code into the then-latest official release. being viewed as a hassle, leaving the player vulnerable; of many systems, entire classes Anyone who downloaded and installed that version during other times these messages are attempts to exploit the media player to install malware on the user’s computer. of vulnerabilities can start to the days it was up on the site ended up making their blog vulnerable to remote PHP execution by online criminals. combine, so that individual In response, WordPress released a new version and hardened its servers. vulnerabilities that may have seemed relatively harmless Different Types of Malware Detected (by month) alone can turn into a serious 90000 risk factor when partnered 80000 with other threats.” 70000 browser help object spyware —Greg Spillman, Cisco Security Analyst rootkit 60000 browser exploit Malware Volume trojan 50000 toolbar hijacker 40000 30000 2008 saw a rise in the use of malware such as Trojans, 20000 browser helper objects and spyware. This data reflects a trend toward more dangerous, 10000 data-gathering malware as well as increasingly clever 0 social engineering vectors. 12/1/07 1/1/08 2/1/08 3/1/08 4/1/08 5/1/08 6/1/08 7/1/08 8/1/08 9/1/08 10/1/08 Date All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 32 Cisco 2008 Annual Security Report

  • Page 35

    Besides these well-known Web-based technologies that DNS Vulnerabilities database it manages, or it can relay the query to other have proven to be vulnerable to attack, the growing crop of DNS servers that manage other portions of the global DNS The big online security concern of 2008 may have been new Web 2.0 technologies such as widgets and add-ons for database. To speed up their response time, DNS servers a Web ecosystem vulnerability that received extensive blogs and social networking sites may also be vulnerable. locally store responses they receive from other DNS news coverage at the end of the summer. It involved servers in a local cache for a certain amount of time. There is also the risk that not all of these add-ons are vulnerabilities in a critical part of the Web’s infrastructure, well-intentioned. Developers have already been creating the Domain Name System (DNS) protocol. In an attempt to evade being blocked by IP address malware distribution, management, and support packages. blacklists, botnet operators and other online criminals often The function of the DNS protocol is to resolve URLs and Social engineering continues to be widely used by online take advantage of the DNS server’s lack of restrictions hostnames such as “cisco.com” to their numerical IP criminals, many of whom have become aware of the on how frequently the records of a domain name and addresses, or IP addresses to URLs. The DNS value of social networks. Therefore, it seems logical that its associated IP address can be changed. Every few protocol allows users to find websites by typing in some of these developers would turn their attention to minutes, the malsite operators transfer the task of hosting “http://www.cisco.com” rather than “”. creating custom, highly appealing “mal-widgets” for social a malsite from one botnet node to another. This practice This makes it easy to associate domains and related networking sites. is called “fast-flux,” or domain-name kiting. subdomains with each other, even if the servers they are hosted on are not physically near each other. Worse, the DNS protocol itself—not just the lack of ActiveX Vulnerabilities restrictions around changing the records in the DNS— Vulnerabilities in ActiveX controls, which power many DNS servers keep records of which domain names go has been shown to have exploitable vulnerabilities in Microsoft applications and Windows applications, with which IP addresses. When a DNS server receives a the area of “cache poisoning.” including the widely used Internet Explorer Web browser, request to resolve a domain name in an IP address from continue to appear in very large numbers. Exploiting these a DNS client, it can look into the portion of the global DNS vulnerabilities typically involves convincing a user to visit a malicious website that invokes a vulnerable ActiveX control. Attackers used vulnerabilities in the Microsoft Snapshot Annual Urgency Scores Annual Severity Scores Viewer, RealNetworks RealPlayer, Microsoft Help Visuals, and Computer Associates BrightStor ARCserve Backup 100 2000 ActiveX controls to conduct high-profile attacks in 2008. 2006 80 2007 ActiveX vulnerabilities have also been used to propagate 2006 2008 2007 1500 malware such as that which targeted outdated versions of 2008 60 RealNetworks RealPlayer for Windows. As evidenced by its success, users often have outdated versions of ActiveX 1000 40 controls installed. As in many situations, even though the vendor released an update to resolve the vulnerability, 500 20 many users hadn’t updated the software. 0 0 Urgency 3 Urgency 4 Urgency 5 Severity 3 Severity 4 Severity 5 These graphs show that both the urgency of vulnerabilities and threats (which is the equivalent of activity) and the severity (equal to the impact) are continuing to increase. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 33

  • Page 36

    Recent Attacks Using DNS Cache Poisoning The Importance of DNS Almost everything the Internet is used for—not just the Web, but also email, FTP, voice over IP, banking China Netcom. DNS cache poisoning claimed one of AT&T Internet Services. A DNS server resolving DNS transactions, and more—relies on the DNS. The DNS acts China’s largest ISPs, China Netcom, as a victim. The online queries for customers of AT&T Internet Services (formerly as the master map of the Internet. Users assume that map criminals who poisoned the DNS cache worked to stay SBC) in the area of Austin, Texas, was compromised. It is correct, but if criminals can modify copies of that map, low-profile, limiting the URLs they redirected to malicious redirected AT&T Internet Services customers who tried they can send anyone using that copy of the map (the DNS websites to those misspelled by users attempting to visit to visit google.com to a malicious page that showed a cache) to a completely unexpected destination, even while certain legitimate websites. This limited the number of fake version of the Google page and incorporated iFrame the users are being shown that they’re following the map victims, but made the attack sneakier and presumably exploits. slower to detect. to their desired destination. This is an especially crucial issue because the DNS is built to be distributed, with different DNS servers owning and trading different parts of the map. They update their parts of the map on a regular basis, but in between updates they use the stored versions of the map (their DNS caches) to send users querying that system to their destinations and give the stored versions of their part of the map to other DNS servers as guides to their neighborhood. With the system set up as it is, a single point of failure— the compromise of even one DNS server—can allow an attacker to poison the cache and mislead all users querying that DNS server. And some of these DNS servers, for example, those of Internet service providers, serve and can potentially mislead millions of users. DNS Cache Poisoning DNS cache poisoning lets online criminals make a legitimate domain name redirect not to the IP address that domain name is supposed to be affiliated with, but to an IP address of their choice. That means they can control where Web users go, sending them to malicious websites even if these users never clicked on a malicious link and instead carefully typed in legitimate website URLs. This makes cache poisoning perfect for hosting malware or for making phishing sites even more successful. Most “regular” phishing websites don’t use legitimate URLs, but URLs that look very similar to the real URL; for example, a website in which the letter “l” in the All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 34 Cisco 2008 Annual Security Report

  • Page 37

    domain name is replaced with the number “1”—making TCP Stack Table Virtualization Vulnerabilities it hard for the visitor to detect that they are not actually Implementation Vulnerability Corporate environments are widely embracing visiting Mylegitimatebanksite.com, but instead Recently, a security researcher disclosed that both virtualization. Whether virtual or remote workers, virtual My1egitimatebanksite.com. known and unknown weaknesses in the TCP stack table data centers, or network virtualization, all offer benefits Getting visitors to these fake sites usually requires using implementations of many products could be exploited in the areas of cost-effectiveness and flexibility. Data some kind of social engineering technique. But when using an exploit called Sockstress. Detailed research has center and network virtualization as well as “virtual client” criminals poison DNS caches, they don’t need to use not yet been released, but initial findings suggest that products may also enhance ease of administration this subterfuge or send out spam linking to the not-quite- affected products could include most operating systems, and security. legitimate URLs. Instead, they use cached DNS records routers, intrusion prevention systems (IPS), and firewall However, some of these virtualization products are stored on DNS servers to control where Web users who devices, since they all handle TCP traffic with stacks that still relatively immature, and have not been rigorously correctly type in or click on a legitimate URL go. They could be affected. tested for security in live environments. This led to 103 essentially hijack all of the Web users who type in that The researcher is known to be working with vendors and vulnerabilities being exposed in virtual software products URL (not just the smaller percentage that clicks on a link in organizations to assist in creating fixes for the affected between January and November 2008. In that same time an email or on another site), so that instead of ending up TCP stacks. Depending on the time required to develop frame, major virtualization vendor VMWare issued 18 at the legitimate destination they typed in, the users are fixes, full information may not be released until sometime security advisories for its products in 2008, compared to redirected to a malicious site. in 2009, when this vulnerability will undoubtedly receive seven advisories for all of 2007. For instance, typing in the legitimate URL additional attention. Mylegitimatebanksite.com would not lead to that legitimate site, but could instead directly send users to a Networking Equipment Vulnerabilities site that tries to download malware onto their computers, or one that looks similar to the bank’s website but sends Although many IT departments spend significant effort “As virtualization technology patching and upgrading desktop systems, applications, any information or passwords visitors type in straight to and data center equipment, upgrading networking gains in popularity, it may bring online criminals. equipment sometimes gets short shrift. This can be In mid-2008, major headlines were generated about a way because if the network is working well, it doesn’t seem like with it new risks.” of exploiting vulnerabilities in many vendors’ DNS server a good idea to interfere and cause network downtime— —Don Simard, Commercial Solutions Director, software that could make it easier to poison DNS caches. and upgrading networking equipment can be complex. U.S. National Security Agency in InfoWorld, Although DNS cache poisoning is not new, security However, not implementing regular upgrades to March 13, 2008 researcher Dan Kaminsky identified a potentially more networking equipment can be dangerous. The amount of reliable and effective means of doing so. research into vulnerabilities in networking equipment and For more information on DNS best practices, network operating systems increased in 2008, including for Cisco protection, and attack identification, visit www.cisco.com/ products. And if exploits do start showing up in the wild, web/about/security/intelligence/dns-bcp.html. the consequences of attacks on corporate networking equipment could be severe. Unscheduled downtime is one potentially painful consequence. Or worse, sophisticated attackers could leave the network running smoothly, and focus on compromising and gaining access to sensitive data residing all over the network. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 35

  • Page 38

    The increasing use of virtualization technologies in One encryption system known to be weak remains in corporate environments is likely to make them attractive widespread use: WEP, which is used for Wi-Fi networks. targets for additional attacks and exploits in the coming year. WEP was broken years ago, attack and exploit tools are widely available, and hacking into WEP-encrypted Wi-Fi Encryption Vulnerabilities networks is easy. Yet many individuals and organizations The growing number of employees working from remote continue to use WEP, which leaves them vulnerable to locations, the increased risk of data loss through error or criminal activity. malice, and the urgent need to protect important infor- Organizations that process credit card information, such mation make encryption a key security tool. Organizations as merchants and service providers, will soon be required are depending on encryption to secure email communi- (by the 2008 update to the Payment Card Industry Data cations, shared data repositories, and devices such as “The more complex the threats Security Standard) to upgrade from WEP to the stronger laptops, CD-ROMs, flash drives, and other memory WPA encryption for their wireless networks. But the myriad become, the more you have to devices that include sensitive data. organizations that provide free Wi-Fi access are not However, several encryption technologies have shown required to make this switch, leaving the security of their do the basics and groundwork vulnerabilities. And weaknesses in encryption can networks porous. Many home users of Wi-Fi access points really well. Staying aware and cause a false sense of security, with users and network also leave their networks vulnerable to snooping and administrators thinking they are protected from certain exploitation. on top of new vulnerabilities threats when, in fact, they are not. Operating System Vulnerabilities and ensuring that patches and In one high-profile example, certain versions of the open Although vulnerabilities that affected all major versions of source operating systems Debian and Ubuntu contain software updates are rapidly an OpenSSL vulnerability that could lead to pseudo- the Microsoft Windows OS and the Linux kernel showed random values being generated—and that could be easily up in 2008, overall, the number of OS vulnerabilities implemented is crucial.” predicted. Using these values could also generate weak discovered declined compared to previous years. Most encryption keys and certificates or passwords, which of these vulnerabilities require user interaction; very few —Jeff Shipley, are exploitable by unauthenticated remote attackers if the Cisco Intelligence Collection Manager would then be vulnerable to brute-force attacks. At the end of August 2008, it became clear that online criminals were victim does not open a file or otherwise perform a required using stolen SSH keys to attack servers running Linux, and action. installing a malicious rootkit on them. There is speculation Widespread acknowledgement of the importance of that the OpenSSL vulnerability in Debian and Ubuntu may patching and regularly updating operating systems—and have played a role in these attacks. making patching easier—have significantly contributed With many organizations using Linux-based servers— to the decline in OS vulnerabilities. Microsoft’s efforts which this exploit laid open to control by online criminals— in this area have been quite successful. So although to run important parts of their networks, this was an OS vulnerabilities are still being reported, their decline extremely serious concern to many IT departments. indicates that attackers are increasingly looking to other classes of vulnerabilities to compromise systems and user information. 36 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 39

    Vulnerabilities in Databases and Mobile Device Vulnerabilities Office Productivity Applications The BlackBerry, an essential part of modern workplace The use of vulnerabilities in office productivity appli- productivity, suffered from a vulnerability this year that cations to conduct both targeted and widespread attacks could compromise corporate networks. Research In continued in 2008. High-profile malicious code attacks Motion (RIM), the makers of the BlackBerry, disclosed that involved products such as the Microsoft Office suite, the way BlackBerrys open PDF attachments could leave Microsoft Jet Database Engine, Adobe Acrobat, and corporate networks vulnerable to being compromised. Ichitaro word-processing software from Japanese office Other smart phones are vulnerable as well. Weaknesses productivity tools vendor JustSystems. in installer applications have allowed Trojans to be installed Interaction from the victim—for example, opening an on certain phones. And the Web browsers that some attached document or malicious database file—is typically mobile phones use may make it easier for users to fall required for criminals to exploit these vulnerabilities and victim to phishing campaigns. For example, a mobile take control of targeted computers. Once attackers have phone’s Web browser may be configured so that the control of the user’s system, they could bypass certain address bar doesn’t show all of a long URL. With phishing perimeter defenses on a corporate network and launch sites, the first part of the URL often looks legitimate, but additional attacks. the latter part may give clues (an “off” top-level domain, or strings of numbers) of being a phishing site. Or the mobile Files associated with these types of applications are well- phone’s input method may make the process of manually suited for targeted attacks by knowledgeable attackers entering URLs into the address bar arduous, so that using social engineering techniques. For instance, an users are more often tempted to “just click” on a possibly attacker might send a malicious spreadsheet labeled “Profit malicious link. and loss statement for shipping department” to people in an organization’s accounting group. The attacker could As with Web 2.0 technologies, some smart phones offer spoof the origin of the document by using easily found an open application development environment, which information from a corporate website, such as the names means that downloading a new application for the phone and email or physical addresses of executives. This is carries the risk that it might be malware. Or, if it wasn’t especially effective as spreadsheets and other office developed using secure coding practices or thoroughly productivity files are commonly used in organizations for tested before release, it might merely be easily exploitable. legitimate business. That means that users often trust them, and they’re rarely blocked at the network perimeter. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 37

  • Page 40

    Geopolitical and Political Conflicts Spam, malware, and botnets are being used to a greater extent as weapons in geopolitical and political conflicts, as in Estonia in 2007 and Georgia in 2008. It is estimated that this trend will continue in the years to come. 38 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 41

    In the 2007 “Estonian Cyberwar” (said to have been a In many of these cases, it is and will remain very difficult revenge attack in response to the Estonian government’s to prove state backing of DDoS attacks against enemy removal of a statue of a Russian soldier from a prominent websites. However, from a security perspective, being location in the capital), Estonian government, banking, aware that geopolitical conflicts are more likely to include and media websites were attacked and shut down using an Internet component—whether it is state-sponsored, or botnet-based DDoS attacks. Speculation continues actions from individual hackers—can help organizations about whether these online attacks were spontaneous, or prepare for the chance that DDoS attacks may be used occurred with Russian state backing. against a country’s government, financial, media, or vital In July 2008, in the weeks leading up to and during the infrastructure websites. Russian-Georgian conflict, Georgian government websites In an interesting political twist, the 2008 U.S. election were defaced or shut down, as was the website of the also saw DDoS attacks against the websites of certain National Bank of Georgia. According to reports in the political campaigns, such as that on a website urging International Herald Tribune, attacks were hosted out votes and soliciting donations to counter a high-profile of servers in the U.S. as well as Russia, attesting to the proposition banning gay marriage in the state of California. “You could fund an entire flexibility of “cyber-warfare.” Botnets affiliated with the Using a DDoS attack, proponents of the proposition were Russian Business Network, a group of online criminals with temporarily able to deny visitors access to the opposition’s cyber-warfare campaign for the ties to the Russian government, were used in the attack. website and impede their ability to donate money to its campaign during a fundraising drive. cost of replacing a tank tread, The Burmese junta has also used online methods against those protesting its regime. During the 2007 political Awareness that botnet activity is likely to increase during so you would be foolish not to.” protests in Burma, the junta shut down all Internet access geopolitical and political conflicts may also be helpful in for the country. (Burma has only one ISP, owning satellite creating a proactive security strategy. And the apparent —Bill Woodcock, Packet Clearing House in The New York Times, August 13, 2008 phones is forbidden, and computers in Internet cafes log weakness of many state-run networks is important to user activities by automatically taking a screenshot every address. If security professionals at these organizations five minutes.) This attempt to stop protesters from sending remain alert to the fact that their networks and websites out digital photos and reports of the political protest was may become targets during conflicts, they may be able largely unsuccessful, so on the anniversary of the protests to strengthen their networks earlier and more thoroughly. in September 2008, the junta reportedly used DDoS For example, they could proactively monitor online attacks to shut down dissident websites. discussions of techniques that may be used against them, allowing them to counter attacks with patches and workarounds. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 39

  • Page 42

    From Conflicts to CyberCommands For many years now, what could be termed low-level Cyber-warfare has not been well-defined except as an cyber-warfare has existed between semi-organized extension of existing electronic warfare; interception hacker groups with political, religious, and other and exploitation of communications would certainly motivations—for instance, between Israel and Palestine, be included under this concept. For example, in mid- China and Taiwan, India and Pakistan, and others. A recent 2008, unclassified White House emails were exfiltrated, change is the addition of overt state sponsorship and according to FBI sources. Origins of the attacks were military backing for Internet-based warfare. traced back to servers in Russia and China, although the existence of state backing is difficult to prove. Security The escalation of cyber-warfare from semi-organized firms linked to the campaigns speculated publicly that individuals or groups to state-sponsored activities brings a foreign entities may be pursuing a “grains of sand” new level of resources, capabilities, skills, and organization approach, in which large amounts of less-well-protected to this arena. The governments of several countries, data is being carefully sifted for nuggets of important including the U.S. and China, have set about establishing information. “cyber-command” organizations. These organizations are tasked with protecting their respective countries from In this context, it is interesting to note that governments online warfare and with creating offensive cyber-warfare around the world are implementing many privacy capabilities. and wiretapping laws, or granting immunity to the telecommunications firms that enable wiretapping. Even though more countries are pursuing this, there is an Recent examples include a proposed U.K. law that would ongoing debate (often outside of military circles) about allow the government to collect data on all electronic whether or when an offensive cyber-warfare capability is communications. Another U.K. proposal would require warranted or a sound decision. That is the experience of user registration for all mobile phones, allowing the many network administrators has shown that trying to go government to create a central database of all U.K. mobile on the offensive against Internet attackers seldom proved users. Notwithstanding significant popular opposition to to be a sound, responsible, or fruitful decision. these proposals, this indicates an ongoing commitment to more closely monitoring communications that may have security implications. In many countries, average citizens will not be affected by this monitoring, unless there are security implications, such as international phone calls to countries or individuals of concern to state authorities. 40 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 43

    Countering Internet Security Threats Apart from working to minimize vulnerabilities and fighting back against current Internet security threats on a case-by-case basis, several broader initiatives are also being used in the battle for online security. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 41

  • Page 44

    Security vendors and researchers are collaborating more • British hacker Gary McKinnon may be extradited DNSSEC closely on the disclosure of vulnerabilities, so that patches and tried in the United States on charges of hacking and workarounds can be created before the exploitable into NASA as well as U.S. Pentagon, Army, Navy, and Industry and governments are working hard to information is widely available. Security vendors are Air Force computers. McKinnon claimed he never mitigate DNS vulnerabilities—the hot issue of summer also working both together and separately to make it harmed the computers, but was looking for evidence 2008. Implementing Domain Name System Security easier to report and discover current security incidents, of alien technology. Extensions (DNSSEC) is widely seen as crucial to and to assess threats accurately. Government initiatives • A U.S. District Court, acting on information collected by ongoing Internet security, in that DNSSEC will provide designed to enhance security are being implemented in the Federal Trade Commission (FTC), shut down and integrity of DNS information to protect against spoofing several countries. And law enforcement is working to send froze the assets of a major international spam network and cache poisoning exploits. online criminals to jail. known as HerbalKing. HerbalKing sent out billions of Although most experts agree that deploying DNSSEC spam messages to market potentially unsafe versions is crucial, adoption faces several challenges, such Industry and Government Initiatives of prescription drugs. The FTC received more than as implementation complexities and disagreements In two separate incidents, hosting providers for online three million complaints about messages related to over who should own the top-level root keys. The criminals were shut down (InterCage in September 2008 this operation. The court froze the spam network’s top-level country code domains of Sweden, Bulgaria, and McColo in November 2008), thanks to efforts by assets and issued a temporary injunction that prohibits Puerto Rico, and Brazil already use DNSSEC. And U.S. security researchers or organizations, law enforcement, the defendants from sending spam and making false officials recently announced that DNSSEC would be or ICANN. In both cases, the amount of spam sent out product claims. New Zealand authorities, working with implemented for the .gov domain by January 2009, worldwide decreased noticeably for several days, until the the FTC, also took legal action against the spammers, and for all .gov subdomains by December 2009, which hosting providers’ clients found other providers. Although and the U.S, government is planning to pursue should further spur worldwide adoption. the long-term impact was limited, this was a positive step: criminal charges. Industry and law enforcement organizations were able to identify and collect evidence showing the malicious On the industry and government collaboration front, new activity, and more importantly, positive action was taken by organizations such as the Industry Consortium for the higher-level service providers to InterCage and McColo Advancement of Security on the Internet (www.icasi.org) are addressing multi-vendor global security threats and and organizations like ICANN. creating a forum for industry collaboration and innovation There are other recent examples of law enforcement and around security. courts working to stop or prosecute online criminal activity. And industry standards continue to be updated to reflect • In July 2008, Seattle “spam king” Robert Soloway changes in technology and security. The Payment Card was sentenced to 47 months in prison. Notorious for Industry Data Security Standard (PCI DSS), for example, marketing spamming services that used botnets to was updated in 2008 to mandate more secure wireless send billions (or by his own account, trillions) of spam encryption technologies. emails, often with spoofed headers that made it appear Meanwhile, National Computer Emergency Response as though they came from Hotmail or MSN accounts, Teams (CERTs) continue to play valuable roles in assessing Soloway finally pleaded guilty to mail and email fraud. threats and vulnerabilities, providing information on them, and coordinating vendor response, as do other government and industry associations. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 42 Cisco 2008 Annual Security Report

  • Page 45

    In the area of identity theft, the U.S. government’s 2008 Enabling Technologies Making Security Identity Theft Task Force Report indicated that in 2007, 2470 criminals were charged with identity theft-related Security vendors are actively working to make security simpler, which helps to enhance the implementation of Easier to Find crimes, while 1943 were actually convicted. The high and adherence to security tools and policies. rate of conviction for those prosecuted, combined with A useful industry standard was created to make compliance requirements and more organizations following For example, vendors have been disclosing threats and it simple to report security incidents and discover best practices around safeguarding personal information, releasing patches more quickly. And to help with crucial information about current security issues. This is helping to reduce the likelihood of identity theft. user education efforts, many vendors have been making it involved having companies create a standardized easier for end users to find information about security risks high-level “security” page on their websites, with a Yet criminals can still make large profits from identity location that would be easy for visitors to remember. and threats and to assess their potential negative effects. theft-related crimes while the probability of prosecution So, a user wanting to see the latest information about is currently low. (To ensure sufficient prosecution of such For many vendors and users, making security simpler a known vulnerability that affects products from a crimes, the report recommended that the government with technology means creating security solutions that: company known as Example would just need to type review its civil monetary penalty programs.) • Make it easy to establish and adapt security policies “ www.example.com/security” into a browser. Although some companies have implemented this standard, More thoroughly tackling identity-theft-related crimes • Automate security tasks (such as encrypting many are still lagging. will require a comprehensive approach toward ensuring sensitive data or deploying patches) greater individual awareness and additional security • Offer protection within, as well as at the edges of, measures by businesses, as well as prosecution and Standardized Security Page the ever-expanding network international cooperation. To make further progress, • Closely monitor and assess threats in real time cisco.com/security increasing the reporting of such crimes to and cooperating with law enforcement will be especially important. • Protect from threats along multiple vectors microsoft.com/security • Integrate with other security tools adobe.com/security • Empower workers to safely use new collaboration and productivity tools yahoo.com/security facebook.com/security Not Yet apple.com secondlife.com google.com mcafee.com myspace.com All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 43

  • Page 46

    Conclusions Threats that combine one or more online elements—the Data loss continues to be a challenge that can have and Key Web, spam, malware, and botnets—continue to grow in number and sophistication. For greater effectiveness, grave, costly effects on organizations and individuals. Creating strong security policies can help, but these policies must be implemented and enforced throughout Recommendations criminals are more and more often targeting specific individuals or groups and exploiting legitimate websites the organization. Regarding data loss, many organizations and other trusted entities and systems. They are launching now assume that company equipment with sensitive increasingly hard-to-detect threats that can dupe even information is likely to go missing at some point. As a result, savvy, cautious users. And they use social engineering they are also increasingly using tools and technology— techniques and take advantage of current events to make including virtual private networks, content filtering at the their online schemes appear highly credible or appealing gateway, authentication technologies, access controls, to their victims. encryption, and data removal and truncation—to keep Meanwhile, the malware that is spread through online sensitive information from being accessed or used by threats is constantly being redesigned to be smarter unauthorized persons. and more surreptitious than ever before—and it has Insider threats are another issue that requires awareness an enormous growth rate. Botnets, the core of criminal and vigilance. In a troubled global economy, more of these activity on the Internet, continue to spread malware, send attacks can be expected. out millions of spam emails, host malicious websites, On the upside, expect to see more news stories in 2009 and attack legitimate ones. To counter these threats, related to how authorities are combating offenders and organizations should include active anti-malware and spammers. By offering specialized, complex services, botnet prevention components in their security strategies. modern offenders are becoming more established, and Online criminals are exploiting both old and new tracking and catching them may become easier. However, 2008 marked both an expansion and weaknesses in technologies and systems to create botnet even as security vendors and authorities collaborate to evolution of the online security threat armies. Known high-impact vulnerabilities are going bring online criminals to justice, this does little to diminish landscape. In some cases, online unpatched. At the same time, increasing use of mobile the number of attacks. devices and remote working, Web 2.0 tools, virtualization, criminals reaped rewards of tens and new forms of collaboration are all expanding the or hundreds of millions of dollars. security perimeter and making the edges of the network This potential for profit will continue more permeable. This poses a significant challenge when organizations try to shore up their defenses, and to drive nefarious innovation and underscores the need for adoption of advanced security specialization in the year ahead. policies and technologies. 44 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • Page 47

    Putting IT on the Front Lines But, if IT is to be more involved and effective at helping to ensure security across the enterprise, organizations According to Cisco’s August 2008 InsightExpress report, must invest more in their IT departments. Ensuring that The Challenge of Data Leakage, approximately 50 percent IT departments can access adequate resources and the of business workers worldwide mix business and personal most knowledgeable and experienced professionals— use on their computers. For most companies, the blending particularly, security specialists —is key. of business and personal IT use is inevitable, but it makes defining and enforcing acceptable use policies that much It is also important to change the perception of the IT more challenging. department’s security policies from “forbidding” to “empowering.” For example, when IT personnel recognize In fact, a recent study by Cisco of the common failures of that employees will download the tools they want enterprise security policies revealed gaps between IT’s regardless of IT policies, they can offer realistic solutions perceptions of why policies are violated, and employees’ to the problem, such as providing vetted versions of true motivations. Most employees surveyed said they downloadable tools and creating secure pathways to broke security policies because the policies either did content on work-related Web destinations. not align with the realities of their jobs, or they needed to access applications not included in the policy, or When employees do inevitably make a mistake or both. Yet most IT professionals thought apathy and lack inadvertently download something that compromises of awareness were the typical reasons for employees’ security, they should be encouraged to be open about security policy violations. it with IT so the issue can be addressed quickly. If the incident is the result of simple human error, without Many companies are struggling to define acceptable use malicious intent, organizations should take a stance of policies that enhance security, but are not so inflexible that demonstrating appreciation to the employee for helping they stifle collaboration and the art of getting business to swiftly identify and combat the threat. This will done in today’s highly competitive, Web-enabled world. encourage users not to feel fearful about informing the IT personnel can help with this, and should be at the IT department about risks. forefront of combating security risks. According to the Employees can play a vital role in safeguarding their 2008 Global Information Security Workforce Study from own online identity and understanding the risks that go Frost & Sullivan, qualified and experienced personnel hand-in-hand with their use of technology. Ongoing user are the key to stopping security threats. They can work education around security policies and technologies and directly with management and employees to create and online threats can help. An important benefit of companies implement relevant and user-friendly policies that are educating and training their workforce about security risks practiced throughout all levels of the organization, starting and threats: It can ultimately lead to better technology in the boardroom. practices in the employees’ personal lives, thereby helping to keep online criminals at bay on two fronts. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 45

  • Page 48

    Key Recommendations Patch known vulnerabilities. To be effective in today’s landscape of evolving threats, security organizations Stay focused. One essential thing for organizations must be aware of new trends. However, many current and to keep in mind about security is that when they try to emerging threats take advantage of known vulnerabilities, protect everything, nothing will be protected. Instead, so organizations should not become wholly distracted by organizations should focus most of their time, energy, Key and resources on what is strategically, financially, and emerging threats. Spending time, energy, and resources on addressing and patching existing defects in their Recommendations competitively most important to safeguard. security armor remains essential. Stop users from inadvertently downloading Checklist malware onto the network. To ensure that users cannot Around 80 percent of common attacks take advantage of 20 percent of high-impact vulnerabilities. These high-value visit—or download the compromised parts of—webpages vulnerabilities are often very basic, and continue to be that contain malware, use several malware scanning unpatched in certain environments. Staying up-to-date 3 Stay focused. technologies; proactive, real-time, reputation-based on these high-value vulnerabilities (and securing them) filtering solutions; and rule- and application-based firewalls will, in most cases, lead to a good “80/20” solution. 3 Stop users from inadvertently and intrusion detection and prevention software at the downloading malware onto the network gateway. Prevent data loss. Strong security policies are essential for protecting an organization from the negative effects network. Individual users can also significantly reduce their chances of data loss. But these policies must be enforced to be of falling victim to malware downloads. Keeping browsers effective, and users must be made aware of them. 3 Patch known vulnerabilities. fully updated and patched, using security features and settings, and remaining aware of existing and emerging Recommendations for reducing the risks associated with 3 Prevent data loss. threats is crucial, as is never clicking on a link received in data loss include: email—even email from apparently trustworthy sources. • Deploy methods (preferably automated) to maintain the 3 Take insider threats seriously. Instead, users should always manually type in a trusted confidentiality of information on mobile devices such as URL, bookmark it, and revisit it by using the bookmark laptops, thumb drives, and PDAs through methods such 3 Remember the network. rather than by clicking on links provided by others. as access controls, encryption, remote data removal, data association, redaction, truncation, or other methods 3 Think beyond compliance. that effectively render data unusable. • Classify data and put stronger controls on what 3 Make security simpler. data people can access. • Define which data should be protected, so that the focus is on keeping the most critical information the most secure. • Educate users about what information should not be stored on a laptop or other mobile device, and what to do if such equipment is stolen. • Actively monitor email and Web traffic to ensure that sensitive information is not being shared inappropriately. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. 46 Cisco 2008 Annual Security Report

  • Page 49

    Take insider threats seriously. Be vigilant by continually Make security simpler. Above all, make security tools logging, auditing, and monitoring traffic patterns, systems, and solutions easier to implement and use, and make and databases. Set policies that prevent employees from security policies easier to follow. engaging in unauthorized activities. Businesses should • Layer integrated sets of security technologies, rather ensure that their information security teams coordinate than depending on patchwork and point solutions. with physical security teams and HR departments to • Ensure that security solutions and departments implement effective policies for revoking access of effectively share data. terminated or transferred employees. Always conduct thorough background checks during the hiring process. • Set security policies that protect all important assets, and make their implementation automated Remember the network. Enterprise security is not and straightforward. just about headline-grabbing malware threats and data • Teach IT departments and employees to work breaches. Despite being a security blind spot for many organizations, network devices are at risk, too. together to enable safe access to productivity- Many organizations use an “if it’s not broken, don’t fix it” enhancing tools and content—both within and approach to their networks—especially in a cost- outside corporate networks. conscious, down economy. While most organizations • Keep existing network and security hardware and should upgrade their networks at least once a year, the software patched and updated in ways that don’t optimal upgrade frequency depends on size, complexity, impede productivity. security requirements, resource constraints, and • Work continuously on user education and aware- other considerations. ness of new threats, and encourage users to report Think beyond compliance. Organizations should not possible or suspected gaps in security. let compliance be a security placebo. By focusing almost exclusively on compliance, and aligning their procedures to meet those requirements, organizations can lose sight of the current, rapidly evolving risk and threat environment. In addition to reviewing compliance levels, organizations should conduct regular top-down gap analyses to augment existing procedures and proactively meet current and emerging threats. All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information. Cisco 2008 Annual Security Report 47

  • Page 50

    Top Trends to Smaller, More Frequent, Targeted Attacks More sophisticated attacks will occur in the year ahead. Reputation Hijacking Hijacking reputations has proven attractive and effective Expect in 2009 They will be deployed rapidly and designed for even more specific targets—individuals, groups, businesses, for online criminals. When people trust a brand, they are likely to visit an associated site or open an email from that organizations, and governments. The current worldwide source without question. Many traditional or point security financial crisis is still playing out, natural disasters and solutions depend on URL or IP filtering lists and don’t manmade strife will continue to provide global news hooks, have real-time insight into traffic patterns and suspicious and a new U.S. president is taking office in 2009. Criminals behavior from every element on a webpage; these will certainly keep refining how they take advantage of (and solutions are not equipped to recognize that a trusted profit from) these types of news events. website or email sender has gone bad. Social engineering and phishing techniques have been In 2009, more online criminals will be actively hijacking profitable, so offenders can be expected to keep refining reputations and will work on finding additional, more the delivery method for (and improving the success of) sophisticated ways to do so. these attacks. There will be more “specialists”—criminals who deliver one or more key components essential to Mobility, Remote Working, creating a complex and convincing attack. As they grow and New Tools as Risk Factors their expertise and reputation, these specialists will be The trend of remote working and related use of Web-based sought out and hired by others looking to create their own tools, mobile devices, virtualization, “cloud computing,” and high-impact attacks. similar technologies to enhance productivity—especially in an economic climate that demands leaner, more-cost To help organizations develop their Cross-Protocol Attacks effective and global staff—will continue in 2009. security strategies and plan their Online criminals looking to improve their odds of success This means that preventing loss of data—from outside will increasingly rely on cross-protocol or “blended” attacks, insiders, or negligence around data storage IT budgets for 2009, Cisco has approaches that combine email, Web-based threats, devices such as laptops—will become more crucial than identified the following key trends to and intrusions. This type of attack, successful in recent ever. But it will be a challenge for security personnel. years, will keep growing during 2009. Also expect to see watch for in the year ahead. These The edge of the network is expanding rapidly, and the more botnets that are capable of “multitasking”—for increasing number of devices and applications in use predictions are based on news and instance, sending spam, hosting malware, and launching make the expanding network more porous, creating new events from 2008, as well as related a direct attack. inroads for threats. information and insight provided To defend against more robust multi-protocol attacks, Organizations of all types should implement thorough, organizations will need to implement security systems that sensible data loss prevention (DLP) policies and consider by Cisco’s security and business can monitor all Internet traffic types and rapidly identify and security solutions that automatically prevent sensitive data operations worldwide. stop new threats. Security solutions that focus on only one from leaving protected environments. area (such as email, IPS or Web-based threats), or those that cannot effectively correlate data between areas, will not Every organization should also begin to take simple steps be enough to protect organizations from blended threats. designed specifically to protect intellectual property—an increasingly precious asset in the modern economy. 48 Cisco 2008 Annual Security Report All contents are Copyright © 1992–2008 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.

  • View More

Get the full picture and Receive alerts on lawsuits, news articles, publications and more!