Malware is becoming more pervasive and much more sophisticated, with the malware "industry" continuing to thrive in the first half of 2007.
So far for the year, the IBM Internet Security Systems (ISS) X-Force research and development team has identified and analyzed more than 210 000 new malware samples, already exceeding the total number of malware samples observed over the entirety of 2006.
X-Force uncovers in the report that the "exploits as a service" industry continues to thrive in 2007.
The 2006 X-Force report indicated that managed exploit providers had begun to purchase exploit code from the underground, encrypt it so that it could not be pirated, and then sell it for top dollar to spam distributors. In 2007, these exploit providers have added the new practice of "exploit leasing" to their repertoire. By leasing an exploit, attackers can now test exploitation techniques with a smaller initial investment, making this underground market an even more attractive option for malicious perpetrators.
According to the report, Trojans (seemingly legitimate files that are actually malware) comprise the most voluminous category of malware so far in 2007, accounting for 28% of all malware, in contrast to 2006 when Downloaders was the most common category. A Downloader is a low-profile piece of malware that installs itself so that it can later download and install a more sophisticated malware agent.
"The X-Force security statistics report for 2006 predicted a continued rise in the sophistication of targeted, profit-motivated cyber attacks," says Kris Lamb, director of X-Force for IBM Internet Security Systems. "This directly correlates to the rise in popularity of Trojans that we are witnessing this year, as Trojans are often used by attackers to launch sustained, targeted attacks."
The use of Web exploit obfuscation continues to rise in 2007 in an attempt to make it difficult for signature-based intrusion detection and prevention products to detect attacks. In 2006, X-Force data reported that approximately 50% of Web sites hosting exploit material designed to infect browsers were obfuscating, or camouflaging, their attack. In the first half of 2007, that number reached 80%.
Counter to historical trends, X-Force reports a slight decrease in the overall number of vulnerabilities uncovered in the first half of 2007 versus the first half of 2006. A total of 3 273 vulnerabilities were identified in the first half of this year, marking a decrease of 3,3% compared to the first half of 2006. This is the first time that vulnerability disclosure numbers have decreased in the first half of the year in the history of the X-Force vulnerability database, which was developed in 1997. However, the percentage of high impact vulnerabilities has gone up since 2006 from 16% to 21% for the first half of 2007.
X-Force points to several trends to explain the decrease in vulnerability disclosures in the first half of 2007 versus the exponential vulnerability growth trends observed in previous years. First, as the monetization of vulnerabilities and exploits has gained attention and maturation in the underground marketplace, a larger percentage of vulnerabilities are remaining undisclosed and are instead being used covertly for monetary and criminal gain.
Second, the increased use of fuzzing by vulnerability researchers over the last two years has uncovered many of the easier to find vulnerabilities. Fuzzing is a testing technique through which random data is supplied to a software program to try to get it to fail and therefore detect vulnerabilities.
"As more technologies and software get exposed to fuzzing and automated bug finding tools, the industry begins to reach a saturation point in the discovery of these types of vulnerabilities, ultimately contributing to the decrease in overall vulnerability disclosures," says Lamb.
Finally, the number of common coding mistakes and bugs is decreasing as a result of software and technology vendors adopting more secure software development lifecycles and more prudent secure coding practices.
A similarly unexpected trend in this report is that, for the first time, spam message size decreased instead of continuing on a linear growth pattern. This decrease corresponds with a decrease in image-based spam. Since mid-2005, image-based spam has been one of the biggest anti-spam challenges, but in the first half of 2007, the percentage of image-based spam declined to the level of mid-2006, at just over 30%. At the end of 2006, image-based spam accounted for more than 40%of spam messages.
"The decrease in spam message size and image-based spam is a result of spammers adopting and experimenting with newer techniques, such as PDF- and Excel-based spam, as a means to more successfully evade detection by anti-spam technologies," says Lamb.