Facebook's URL scanner is vulnerable to cloaking attacks

Link previews don't always match the reality, even after Websense checks

Members of a hacking think-tank called Blackhat Academy claim that Facebook's URL scanning systems can be tricked into thinking malicious pages are clean by using simple content cloaking techniques.

Such attacks involve Web pages filtering out requests that come from specific clients and feeding them content that is different from what is displayed to regular users.

[ The Web browser is your portal to the world -- as well as the conduit that lets in many security threats. InfoWorld's expert contributors show you how to secure your Web browsers in this "Web Browser Security Deep Dive" PDF guide. ]

Attackers have been using this method to poison search results on Google for years now by serving keyword-filled pages to its indexing robot, but redirecting visitors to malware when they click on the links. However, it turns out that Facebook is also vulnerable to this type of content forging. "Hatter," one of the Blackhat Academy members, provided a live demonstration, which involved posting the URL to a JPEG file on a wall. Facebook crawled the URL and added a thumbnail image to the wall post, however, clicking on its corresponding link actually redirected users to YouTube. This happened because the destination page was able to identify Facebook's original request and served a JPEG file.

"While most major sites that allow link submission are vulnerable to this method, sites including Websense, Google+, and Facebook make the requests easily identifiable," the Blackhat Academy hackers said.

"These sites send an initial request to the link in order to store a mirror thumbnail of the image, or a snapshot of the website being linked to. In doing so, many use a custom user agent or have IP addresses that resolve to a consistent domain name," they explained.

Earlier this week, Facebook signed a partnership with Websense to use the security vendor's cloud-based, real-time Web scanner for malicious URL detection. Blackhat Academy has now provided proof-of-concept code that, according to its advisory, can be used to bypass it.

Websense doesn't believe that to be the case. "This is nothing new. We use numerous methodologies and systems to ensure that our analysis of content (in real time) is not manipulated by malware authors, including using IP addresses not attributable to Websense so that malware authors are unaware that it is Websense analyzing the content," the company said.

"Also, the Websense ThreatSeeker Network is fed via an opt-in feedback loop from tens of thousands of customers distributed globally. These IPs are also not attributable to Websense.com. It is because of technologies like this that Facebook chose Websense to provide protection for their growing user base of more than 750 million users," it added.

That could well be true, but it's worth keeping in mind that Websense primarily sells security solutions to businesses and Facebook is usually blocked on many corporate networks. It would be logical to assume that relying on its customers' appliances to scan URLs on the social networking website might not have an immediate impact.

Hatter says that as a security research outfit Blackhat Academy follows responsible disclosure and notified Facebook of the content cloaking issue at the end of July. Despite this, the method still works.

"We're well aware of the content forgery technique described and have built protections into our systems to account for it," a Facebook spokesman said via email.

"The content returned when we crawl a shared link is only one of many signals we use to combat spam and abuse on Facebook. We know that this content can change between visits, and therefore can't always be trusted, and our systems account for that," he added.

Earlier this year, Facebook signed a partnership with Web of Trust, an organization that maintains a community-driven spam URL block list. However, it's well-known that blacklisting is not very efficient, and there can be a significant window of exposure between the time when a URL starts being spammed and the time when it's flagged by such a system.

At the very least, content cloaking can be a powerful social engineering technique. A link with a .jpg termination accompanied by a thumbnail can look harmless enough to trick a lot of users into clicking on it.

Facebook and Websense are not the only ones with this problem. Google+ and Digg are also vulnerable to cloaking attacks, but other sites, such as Twitter, have developed strong protections against them.