A security researcher has devised an attack that remotely steals digital credentials used to access user accounts on Facebook and other websites by exploiting a flaw in Microsoft's Internet Explorer browser.
Independent researcher Rosario Valotta demonstrated his “cookiejacking” proof of concept last week at the Hack in the Box security conference in Amsterdam. It exploits a flaw that's present in all current versions of IE to steal session cookies that Facebook and other websites issue once a user has entered a valid password and corresponding user name. The cookie acts as a digital credential that allows the user to access a specific account.
The proof of concept code specifically targets cookies issued by Facebook, Twitter and Google Mail, but Valotta said the technique can be used on virtually any website and affects all versions of Windows.
“You can steal any cookie,” he told The Register. “There is a huge customer base affected (any IE, any Win version).”
The attack exploits a vulnerability in the IE security zones feature that allows users to segregate trustworthy websites from those they don't know or don't ever want to access. By embedding a special iframe tag in a malicious website, an attacker can circumvent this cross zone interaction and cause the browser to expose cookies stored on the victim's computer.
The exploit requires the attacker to accomplish a variety of difficult tasks, including knowing where on a hard drive cookies are stored (it can be slightly different for various versions of Windows) and knowing the victim's Windows username.
Valotta's exploit incorporates techniques developed by researchers including one by Jorge Medina that manipulates file-sharing functionality built into IE to transmit the Windows username in plain text. It also borrows an advanced form of clickjacking, known as drag & drop content extraction, which was demonstrated last year by Paul Stone.
A video of Valotta's attack in action is below.
“It is complicated for the attacker but not for the victim,” Valotta said.
Microsoft spokesman Pete Voss issued a statement that read in part:
"We are aware of an issue that could enable theft of a user's cookies if they were convinced to visit a malicious website and once there, further convinced to click and drag items around on the page. Given the level of required user interaction, this issue is not one we consider high risk in the way a remote code execution would possibly be to users."
He also said attackers can steal cookies only from websites the victim is logged into. He went on to say users should avoid suspicious links.
Valotta said he first alerted members of Microsoft's security team in January. He said the company intends to patch the vulnerability in a series of updates tentatively scheduled for June and August. ®