Web-Borne Malware
Web-borne malware exploits a blind spot in many security implementations. Specifically, many applications fail to inspect the data returned from a visit to an Internet Website. Typically using Java Script, malicious hackers literally append malicious code to the Internet Web page. When the unsuspecting user visits the infected Web page, the JavaScript runs in the user’s browser and in most cases causes malware (such as a key logger or root kit) to be downloaded to the user’s PC. This methodology has also been referred to as “Drive By Hacking” and has impacted several high-profile Websites. One prime example: at the peak of the NFL playoffs, hackers compromised the Miami Dolphins’ Website and malware was automatically loaded onto visitors’ PCs. Similarly, the Center for Disease Control Website was compromised, and unsuspecting visitors had malware loaded onto their PCs.

The most notable Web-borne malware event to date involved the storm worm and its respective spam botnet. Starting in January 2007, the storm worm has gone on to impact hundreds of thousands of users. The malware generated over 9 million e-mails to users. Each message directed recipients to Websites containing malicious code. While the social engineering aspects change regularly, the Web 2.0 / Web-borne malware delivery methodology remains the vehicle of choice for the delivery of its root kit / spam bot malware.

Still in Denial
Even today, Web server administrators irresponsibly dismiss Web-borne malware as non-malicious attacks, while thousands of their Website visitors are put at risk. An estimated 450,000 URLs point to malicious Websites hosting malware on the Internet. That’s nearly half-a-million hidden landmines that casual users and even advanced users can’t spot on their own.

RSS / ATOM Feeds
Nearly 12% of Internet-connected users take advantage of RSS / Atom feeds to receive timely news and data content. Alas, most users don’t consider the security ramifications of connecting to a remote server for an RSS / Atom feed. To be sure, using RSS to deliver malware is well within the realm of possibilities.

The RSS threats aren’t new. As far back as 1995, Yahoo was alerted to an RSS feed vulnerability in the company’s RSS aggregator. A more recent security issue has been found with the AOL ICQ Toolbar (CVE-2006-4660). In this example, the default options2.html Web page is not validated before loading. This permits a hacker to execute arbitrary script / coding by tricking them into visiting a malicious Website. A second vulnerability in AOL ICQ Toolbar (CVE-2006-4661) takes advantage of a failure to validate the title and description fields of the feed. This sets the stage for a hacker to trick a user into visiting a malicious Web site to execute arbitrary HTML / scripting.

XSS Scripting
XSS allows malicious hackers to inject code into the Web pages viewed by others. This methodology has become popular in Web 2.0 phishing exploits and browser vulnerabilities. In some cases, it allows the hacker to bypass access controls within the user’s network

There have been several XSS exploits on the Internet and two received a great deal of media attention:

• MySpace XSS Worm: A worm written to exploit an XSS vulnerability in MySpace brought the service down for nearly two days. The exploit injected JavaScript into users’ pages and when the Web page was visited, the JavaScript was executed in the visitors’ browsers.

• Yahoo XSS: XSS vulnerability in Yahoo tricked users to click on a booby-trapped link. After each user clicked on the link, the hacker gained access to the user’s Yahoo account including e-mail, address book, and calendar entries.

It should also be noted that a security researcher recently uncovered 40 flaws in Google’s YouTube Website. The vast majority of these flaws were XSS issues that put users at risk of having their profiles infected with a fast-spreading worm that could potentially steal users’ credentials.

Cross-Site Request Forgeries (CSRF)
CSRF is often confused with XSS exploits. However, CSRF does not rely on any client-side active scripting. Instead, CSRF exploits a victim’s prior relationship and authority with a Website to allow unwanted or unapproved actions by a malicious hacker.

Several CSRF exploits have been reported. One particular CSRF vulnerability in phpMyAdmin (CVE-2006-5116) could allow a remote attacker to perform unauthorized activities posing as another user simply by setting a token in a specially crafted URL. Another CSRF issue demonstrated at the 2006 BlackHat conference showed that a user’s DSL router configuration could easily be changed because the user’s browser automatically supplied the DSL router with the user’s credentials – even though the hacker (not the user) was accessing the DSL router.

CSRF exploits in security products are not limited to DSL routers. In an article at Calyptix, CSRF exploits were found to exist in at least four mainstream security appliances, allowing a remote user to gain administrative access.

XSS Filter Bypassing
With the large number of XSS issues plaguing the Web 2.0 world, administrators began utilizing filters to ward off XSS attacks. Filtering to block XSS attacks can be effective but is often found to be weak when the Website must support multi-national users. Simply put, most XSS filters are written to support a given character set for a specific language. To bypass the XSS filters, hackers simply switched the default language encoding (e.g., UTF-8 to US-ASCII).

XSS filter bypassing issues have been found in several software products including PHPNuke and the firewall product from NetGear (FVS318) that allowed hackers to run XSS attacks against FVS318 administrators.

Exponential XSS Attacks
While a traditional XSS attack targets an individual user, an exponential XSS attack targets hundreds or perhaps thousands of users. Proof of Concept code for an exponential XSS attack can now be found on the Internet.

The example attack, known as the “Nduja Connection,” is a worm that targets Web-based e-mail systems. It collects the e-mail of a given user, collects all of the user’s contact e-mail addresses, and then self-propagates to all of the user’s contacts. This exponential XSS attack was used successfully against Liberi.it, Tiscali.it, Licos.it, and Excite.com Web-based e-mail systems.