Google Disputes Socially Engineered Malware Study

How FiresheepCan Hijack Web Sessions

In the wake of a report by an independent security testing lab that awarded top marks to Microsoft Internet Explorer 9 for its ability to stop socially engineered malware, Google fired back, alleging that such malware only accounts for 2% of malicious code found online.

That statistic comes via a new report from Google, "Trends in Circumventing Web-Malware Detection," that was compiled using four years' worth of data, gathered as part of its Safe Browsing initiative. According to Google, its Safe Browsing API, which provides a block list of suspected or known-bad URLs linked to malware and phishing attacks, results in about 3 million malware warnings daily to more than 400 million people. Firefox, Safari, and Chrome all make use of the data feed.

Google's publicizing of the report followed the release, earlier this week, of a study from independent security testing organization NSS Labs, which tested five browsers for their ability to stop socially engineered malware found in the wild. It awarded top marks to IE9, which it said successfully and quickly blocked almost 100% of all such threats, followed by Chrome (13.2%), Safari and Firefox (7.6%), and Opera (6.1%).

The Google researchers acknowledged that socially engineered malware is dangerous, and that its volume continues to increase. "Social engineering is a malware distribution mechanism that relies on tricking a user into installing malware. Typically, the malware is disguised as an antivirus product or browser plug-in," said two of the Google report's co-authors, Lucas Ballard and Niels Provos, part of the Google Security Team, in a blog post.

Even so, they said, "it's important to keep this growth in perspective--sites that rely on social engineering comprise only 2% of all sites that distribute malware."

In fact, according to their research, drive-by downloads are a much more common attack vector. In such exploits, "malicious pages install malware after exploiting a vulnerability in the browser or a plug-in," they said. Furthermore, attackers often exploit the latest vulnerabilities to help their attacks avoid detection. That said, "a prominent exception is the MDAC vulnerability which is present in most exploit kits," according to the researchers. (MDAC is a set of exploits that target ActiveX controls.)

NSS Labs appears to have predicted Google's criticism. "NSS Labs commends Google for adding some protection against socially engineered malware to Chrome," according to its report. "Further, we view the addition of this protection as evidence that even as Google's PR engine downplayed the findings of prior NSS Labs test reports, the Chrome engineering team was working hard to address this known deficiency."

While blocking socially engineered malware is important, the Google researchers emphasized that stopping Internet-borne malware in general is extremely difficult, owing to the many different ways in which computers can become infected. "The attack surface of the modern Web browser is quite large," said the Google researchers. "Web-based malware can target vulnerabilities in the browser itself, or against the myriad of plug-ins that extend the browser to handle, for example, Flash, Java applets, or PDF files. A vulnerability in any of these components may be leveraged to compromise the browser and the underlying operating system." Furthermore, malicious files can be introduced not just via compromised websites, but also through email and instant messages.

The Google research further highlights that no one defense mechanism alone will stop all malware attacks. In particular, the paper studied the four most prevalent defenses: "virtual machine client honeypots, browser emulator client honeypots, classification based on domain reputation, and antivirus engines."

According to the researchers, "our results show that none of these systems are effective in isolation." In addition, what works today may fail tomorrow, as there's an "arms race" between browser makers and attackers, who are "highly motivated and quickly adapt to technologies that try to protect users from their sites," they said.

At a full-day virtual event, InformationWeek and Dark Reading editors will talk with security experts about the causes and mistakes that lead to security breaches, both from the technology perspective and from the people perspective. It happens Aug. 25. Register now.