Disclosure of Major New Web 'Clickjacking' Threat Gets Deferred

Details of a new major Web attack that could potentially affect millions of users won’t see the light of day next week as planned after the researchers who discovered it agreed to hold off on disclosing their find until Adobe comes up with a patch for its product.

Renowned Web security researchers Robert "RSnake" Hansen and Jeremiah Grossman late yesterday pulled their presentation "New 0-Day Browser Exploits: Clickjacking - yea, this is bad" from the upcoming OWASP USA security conference in New York, after Adobe requested that the researchers give them time to come up with a patch for one of its applications before they release their proof-of-concept code.

Hansen and Grossman just days ago found that a vulnerability that can be used for so-called “clickjacking” attacks wasn’t in Adobe’s application, but in various browsers, including Microsoft’s and Mozilla’s, and affects Adobe's application. It can even evade browser security features. While they can’t give details of the specific vulnerabilities at this time, they say this new clickjacking attack -- where a bad guy lures a victim to click onto a link -- could leverage other Web attacks like cross-site scripting (XSS), SQL injection, and cross-site request forgery (CSRF), to attack a wider range of users.

“It surprised us that Adobe took ownership over an attack technique that we considered to be the responsibility of the browser vendors,” says Grossman, who also blogged on the decision to drop the OWASP (Open Web Application Security Project) talk. “They want to protect their users as best they can no matter what. So when that happened, we had to put the disclosure brakes on.”

OWASP member and researcher Joshua Perrymon says the newly discovered attack lets the bad guy take control of the victim’s audio, microphone, and Webcam and interact with the desktop. Grossman and Hansen, however, wouldn’t comment on those details.

“It lets them own any PC remotely that’s running Flash/IE browser, just by having a user visit a site with the malicious .swf,” says Perrymon, who has done some research on malicious .swf files in the past.

Perrymon says the attack could originate from an email luring a victim to a malicious link, or even via a legitimate Website that’s been infected and carries an embedded malicious Flash file.

“Just like anything JavaScript malware-related... you visit an attacker-controlled page, or one which [has] their code on it like a MySpace/Facebook profile, Message Board, Banner Advertisement, etc.," says Grossman, who is CTO and founder of WhiteHat Security.

Grossman says that an Adobe fix will take care of about “90 percent” of the problem, but a flaw in Internet Explorer 8 would still be outstanding -- but not vital to his and Hansen’s research, he says.

“[Clickjacking is] way worse than we initially gave it credit” for, Grossman says. “This also further reinforced the complete rethink of the browser security model.”

Hansen described the vulnerability this way: “It’s an architectural issue in the browser that affects Websites, plug-ins, and users,” says Hansen, who is founder of SecTheory LLC. He blogged about the aftermath of his and Grossman's decision today: (See Channeling Dan Kaminsky.)

Just where all of this leaves users while patches are under development is unclear. Hansen and Grossman say the best bet is for users to take the unpopular step of disabling scripting and browser plug-ins. “Hopefully Adobe can put out some kind of patch in short order, but who knows the timeline,” Grossman says.

“The problem [is that] the solutions are no good. See, we still don't really think it's their [Adobe’s] bug,” he says.

Whether the bad guys will get there first is uncertain. “Clearly people will speculate about the flaw, and the attack surface area is unfortunately rather large. I still believe there are far easier hacks against browsers out there, but since multiple vendors were affected, we felt obligated to give them [Adobe] time,” Hansen says. “I suppose the greatest threat is that the vector is out there, there are no solid patches for the time being, and people may look deeper into the subject speculatively, a la Dan Kaminksy's DNS bug. That's never a good combination.”

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.