News Application Security
Google Report: How Web Attackers Evade Malware Detection
Data gathered from Google's Safe Browsing API service reveals drive-by infections most common, with 'IP cloaking' on the rise
Attackers increasingly are engaging in "IP cloaking" to infect Web visitors, where they bypass malware detection systems by serving them clean pages while they drop malware on visitors to the site, according to a new report by Google's security team.
"Over the years, we have seen more malicious sites engaging in IP cloaking. To bypass the cloaking defense, we run our scanners in different ways to mimic regular user traffic," said Lucas Ballard and Niels Provos of the Google Security Team, in a blog post yesterday.
More Security Insights
White PapersMore >>
Google's research is based on more than four years of data gathered from its Safe Browsing API service. Google's Safe Browsing API is an online database that contains known malware-rigged Web pages and phishing sites. Chrome, as well as Mozilla's Firefox and Apple's Safari browsers, use the database, as well.
The search engine giant's analysis of the malware-evasion methods used by the bad guys is based on 160 million Web pages hosted on some 8 million sites.
As of summer 2010, 160,000 or so websites were employing cloaking domains, according to Google's report, which covers five years of data. This technique peaked two years ago, when there were some 200,000 websites with IP cloaking, up from more than 50,000 earlier that year. "That peak coincides with a large-scale attack, where thousands of sites were infected to redirect to gumblar.cn, which actively cloaked our scanners," Google says in its report. "Although the increase in the graph is partly due to improved detection of cloaking domains in our system, we believe that it is representative of the general state of cloaking."
Attackers are also using social engineering and drive-by downloads, according to Google. Socially engineered Web attacks try to lure a user into following a link or downloading software. Malware from websites is one of the three main threat vectors for browsers; phishing attacks and exploits are the other two. A recent test by NSS Labs showed that Internet Explorer 9 was best at catching socially engineered malware attacks.
But Google says while social engineering is on the rise as a vector for Web malware, only 2 percent of all websites that distribute malware actually use that vector. The malware often comes in the form of fake antivirus or a browser plug-in, for example.
"Social engineering has increased in frequency significantly and is still rising. However, it’s important to keep this growth in perspective," the Google report says.
Drive-by downloads remain one of the more popular vectors for malware: This is where attackers exploit a bug in a browser or browser plug-in to infect the victim. "Our analysis of which vulnerabilities are actively being exploited over time shows that adversaries quickly switch to new and more reliable exploits to help avoid detection," Google says in its report. "Most vulnerabilities are exploited only for a short period of time until new vulnerabilities become available. A prominent exception is the MDAC vulnerability which is present in most exploit kits."
Google found that the bad guys are constantly retooling their efforts in order to evade the more common detection methods of virtual machine honeypots, browser emulation honeypots, domain reputation, and antivirus engines. "Our experiments corroborate our hypothesis that malware authors continue to pursue delivery mechanisms that can confuse different malware detection systems," the report says.
The full report is available here (PDF) for download.
Have a comment on this story? Please click "Add Your Comment" below. If you'd like to contact Dark Reading's editors directly, send us a message.