Phone Number +1-202-802-9399 (US)

Thycotic is now Delinea!

The Lockdown

Thycotic’s Cyber Security Blog

Don’t FREAK—A Zombie Security Flaw from the 90’s?

Written by Thycotic Team

March 10th, 2015

Nearly a third of encrypted websites were vulnerable to yet another SSL-related security flaw – this one dubbed the “FREAK” vulnerability, standing for Factoring RSA Export Keys. The gist of the FREAK attack is that attackers can intercept traffic between client and server – traffic that is supposed to be encrypted with SSL or TLS.

The “FREAK” Flaw

To understand the vulnerability, we’ll put it in context. According to researchers, the United States government placed export restrictions on software that used strong encryption in the 1990s. The policy is no longer required, but back then, software developers had to use small encryption keys which were able to be defeated using brute-force attacks in order to export their products. SSL was then made to be compatible with “export mode,” a feature which allows two parties to communicate using a smaller than usual key size. In order to be compatible with software from this era, SSL never got rid of export mode, meaning it is still usable today, and is a major threat.

How could an attack take advantage of this?

Well, it starts with a man-in-the-middle attack. An attacker intercepts a communication channel between two people and sends both sides a request to use export mode. Now, the encryption key used for the traffic between the two people is significantly weaker, and can be broken by the attacker. This results in a situation where the attacker can sniff traffic between the two as if SSL wasn’t even being used – remember, a password doesn’t protect you if your attacker knows it!

Who’s Vulnerable?

Currently, Apple browsers, the built-in Android browser, and Internet Explorer are reported to be vulnerable to a FREAK attack. Current versions of Google Chrome and Firefox are not considered vulnerable or already have patches. Apple has said a patch is coming within the next week and Google has released a patch to device makers and wireless carriers for them to distribute. In addition, although nearly 1/3 of websites using SSL were reported to be vulnerable, that number is quickly declining as websites are fixed.

In order to fall victim to a FREAK attack, you would have to be visiting a vulnerable site using a vulnerable browser – the odds of this occurring are rapidly decreasing. You would also have to fall victim to a man-in-the-middle attack, which usually only occurs when the attacker is on the same local network. To make sure you’re safe, download any available browser patches and stay away from public wi-fi hotspots (you should avoid these anyway), and the FREAK attack should not bother you.

The 90’s, Coming Back to Bite Us

Retrospectively, researchers and bloggers alike argue that we should look at the discovery of this vulnerability as a past mistake coming back to haunt us. A University of Pennsylvania cryptographer, Nadia Heninger, even said, “This is basically a zombie from the 90’s.” If the US government had never imposed restrictions on the strength of cryptography in exported software, they argue, this problem never would have arisen. Indeed, it is a painful reminder that security and convenience are often measured inversely – larger keys would have made things less convenient to decrypt via brute force, but they would have ensured there was no need for an “export mode” in the first place.

UPDATE: The latest Apple release, iOS 8.2, fixes the “FREAK” flaw in the company’s products. Learn more here.