Vulnerabilities could enable attackers to collect data on Facebook users and friends, Dhanjani says

Tim Wilson, Editor in Chief, Dark Reading, Contributor

January 20, 2010

4 Min Read

Vulnerabilities in the way members authorize the use of third-party applications in Facebook could potentially lead to loss of personal information or even targeted attacks on specific individuals, a security researcher said today.

Nitesh Dhanjani, a well-known security researcher and author of Hacking: The Next Generation, says he has discovered design flaws in Facebook that could allow attackers to collect the personal information of users on the social networking site, and even build profiles of "friends" that might facilitate direct attacks on specific individuals within a company.

The flaws were presented to Facebook in November; Dhanjani has agreed not to release specific code or other details for two weeks while technical staffers at the social networking site continue their efforts to patch the vulnerabilities. Dhanjani says he has begun to speak generally about the problem, without specifics.

The vulnerabilities center around the way Facebook enables users to place third-party applications on their social networking pages, Dhanjani says. In a nutshell, Facebook allows the use of third-party apps within the confines of the site, but only if the user authorizes them. "If you click on a link that requires a third-party application, you see a dialog box, and you have to click 'yes' to authorize it," Dhanjani explains. "Once you authorize its use, all of your information -- your user ID, your friends list, everything -- is shipped to the third party. I'm not sure people really understand what's happening to their data."

Worse, Facebook also has enabled some applications to provide "automatic" authorization, Dhanjani observes. "When the user visits the application from within the Facebook environment, Facebook inserts "a parameter," he states in a report about the vulnerability. "If this parameter is present when the application is rendered, the application is allowed to scour information from the user's profile. The intention in this situation is that if the user clicked on the application [rather than a third-party site that redirects the user], the user has implicitly granted some level of authorization."

Dhanjani calls this automated authorization a "design flaw" in Facebook, but the social networking site has chosen not to comment on this particular concern. "They want users to be able to use the applications more easily, so it's basically a business decision to leave it the way it is," he states.

However, Facebook is responding to Dhanjani's assertion that flaws in these authorization procedures could potentially be exploited to create clickjacking attacks.

"The goal is to write a rogue Facebook application that is rendered when a user visits a malicious third party Website," Dhanjani explains in his report. "If the user already has an established session in Facebook [on another browser tab or window], the third-party site can load the malicious Facebook application in an iFrame to identify the user and steal the user's Facebook information."

Since only part of the actual Facebook site is being displayed in the iFrame, the attacker is essentially executing a "clickjacking" attack, Dhanjani says. The attacker is essentially creating a malicious application that looks like a legitimate app -- and then when the user clicks on the right link, the malware uses Facebook's flawed authorization process to collect all of the user's Facebook data, including information about the user's "friends."

"We've already seen clickjacking work on Facebook, but those attacks were mostly used to spread spam to users and their friends," Dhanjani says. "What's happening in this case is that the attacker is using clickjacking to collect the data of the user, as well as the data on their friends. You could map that data to specific domains, such as users who are in a company and their friends."

Cybercriminals could potentially use such a flaw to collect data on specific individuals, Dhanjani warns. "If you want to install malware on the computer of a user in a particular business unit of [a corporation], for example, that's pretty hard to do with a traditional browser attack. But with this, you can actually target an individual or build a group of individuals that you want to target with a specific piece of malware."

It's hard to tell how dangerous these attacks might be because the severity of targeted attacks can't be measured in numbers of infections or numbers of instances detected, Dhanjani says. "But I would be very surprised if there aren't already [hackers] looking at this vulnerability," he says.

Dhanjani plans to provide more details on the vulnerability, including specifics on code, in about two weeks -- "Hopefully, after Facebook has fixed the problem," he says.

Have a comment on this story? Please click "Discuss" below. If you'd like to contact Dark Reading's editors directly, send us a message.

About the Author(s)

Tim Wilson, Editor in Chief, Dark Reading

Contributor

Tim Wilson is Editor in Chief and co-founder of Dark Reading.com, UBM Tech's online community for information security professionals. He is responsible for managing the site, assigning and editing content, and writing breaking news stories. Wilson has been recognized as one of the top cyber security journalists in the US in voting among his peers, conducted by the SANS Institute. In 2011 he was named one of the 50 Most Powerful Voices in Security by SYS-CON Media.

Keep up with the latest cybersecurity threats, newly discovered vulnerabilities, data breach information, and emerging trends. Delivered daily or weekly right to your email inbox.

You May Also Like


More Insights