About 8% of Android apps are vulnerable to attacks as a result of weak SSL implementations, according to a new computer security study.
Security researchers in Germany analyzed 13,500 free Android apps from Google Play and found that 1,074--about 8%--contain SSL/TLS code that could potentially make them vulnerable to what's known as a Man-in-the-Middle (MITM) attack.
The researchers, from Leibniz University in Hannover and Philipps University in Marburg, describe their findings in a paper titled "Why Eve and Mallory Love Android: An Analysis of Android SSL (In)Security."
SSL/TLS are cryptographic protocols used to secure online communications. As many security researchers have demonstrated, however, the way these protocols are implemented and their reliance on a trusted third-party Certificate Authority leave many applications vulnerable.
[ Read about Samsung's new Chromebook at Samsung Chromebook: Hands-On Review. ]
This is not a new problem--computer security researcher Moxie Marlinspike released software called sslsniff that demonstrated SSL flaws back in 2002--and more holes keep being found. For instance, researchers from Stanford University and the University of Texas at Austin last week presented a paper titled "The Most Dangerous Code in the World: Validating SSL Certificates in Non-Browser Software." It claims to demonstrate that "SSL certiﬁcate validation is completely broken in many security-critical applications and libraries."
In the case of Android apps, the researchers from Germany created a program called MalloDroid to conduct static code analysis on the networking API calls. The program retrieved the http/https URLs from decompiled apps and checked the extracted SSL certificates for validity. It also flagged apps that used custom trust management practices, which may be more permissive than the Android default. The researchers then audited 100 apps manually to look into potential SSL bad practices.
The researchers also examined how the Android apps presented SSL security information to users, to identify poor interface design and possible UI improvements.
In addition to finding 1,074 apps out of 13,500 that accepted all certificates or all hostnames from a certificate--a risky security practice--the researchers through their manual audit found 41 apps out of 100 were vulnerable to MITM attacks.
The vulnerable apps were not identified, but the paper states that three of them have installed bases of between 10 million and 50 million users.
"From these 41 apps, we were able to capture credentials for American Express, Diners Club, Paypal, bank accounts, Facebook, Twitter, Google, Yahoo, Microsoft Live ID, Box, WordPress, remote control servers, arbitrary email accounts, and IBM Sametime, among others," the researchers state in their paper. "We were able to inject virus signatures into an antivirus app to detect arbitrary apps as a virus or disable virus detection completely."
With regard to the way the apps presented security information, the researchers found that half of 754 Android users surveyed failed to correctly judge the security state of a browser session correctly and that 56% had not seen a certificate security warning before and rated the risk presented in the warning as medium to low.
Google declined to comment.
While the researchers acknowledge that "the default Android browser is exemplary in its SSL use," they nonetheless present several recommendations about how to improve Android security, particularly for third-party apps that use customized SSL code.
They suggest that Google should do the following:
--Disallow custom SSL handling in Android.
--Implement HTTPS-Everywhere as part of its communication API.
--Make its Internet permission model more fine-grained so users can choose to avoid apps without SSL.
Attackers are increasingly using a simple method for finding flaws in websites and applications: They Google them. Using Google code search, hackers can identify crucial vulnerabilities in application code strings, providing the entry point they need to break through application security. In our report, Using Google To Find Vulnerabilities In Your IT Environment, we outline methods for using search engines such as Google and Bing to identify vulnerabilities in your applications, systems and services--and to fix them before they can be exploited. (Free registration required.)