New Android vulnerabilities put over a billion devices at risk of remote hacking

The flaws are located in Android's media processing components and can be exploited through maliciously crafted websites

Android's media processing components are riddled with critical vulnerabilities
Martyn Williams/IDGNS

Newly discovered vulnerabilities in the way Android processes media files can allow attackers to compromise devices by tricking users into visiting maliciously crafted Web pages.

The vulnerabilities can lead to remote code execution on almost all devices that run Android, starting with version 1.0 of the OS released in 2008 to the latest 5.1.1, researchers from mobile security firm Zimperium said in a report scheduled to be published Thursday.

The flaws are in the way Android processes the metadata of MP3 audio files and MP4 video files, and they can be exploited when the Android system or another app that relies on Android's media libraries previews such files.

The Zimperium researchers found similar multimedia processing flaws earlier this year in an Android library called Stagefright that could have been exploited by simply sending Android devices a maliciously crafted MMS message.

Those flaws triggered a coordinated patching effort from device manufacturers that Android's lead security engineer, Adrian Ludwig, called the "single largest unified software update in the world." It also contributed to Google, Samsung and LG committing to monthly security updates going forward.

One of the flaws newly discovered by Zimperium is located in a core Android library called libutils and affects almost all devices running Android versions older than 5.0 (Lollipop). The vulnerability can also be exploited in Android Lollipop (5.0 - 5.1.1) by combining it with another bug found in the Stagefright library.

The Zimperium researchers refer to the new attack as Stagefright 2.0 and believe that it affects more than 1 billion devices.

Since the previous attack vector of MMS was closed in newer versions of Google Hangouts and other messaging apps after the previous Stagefright flaws were found, the most straight-forward exploitation method for the latest vulnerabilities is through Web browsers, the Zimperium researchers said.

Attackers could trick users into visiting websites that exploit the flaw through links in email and instant messages or through malicious advertisements displayed on legitimate websites.

Man-in-the-middle attackers who are in a position to intercept users' Internet connections, for instance on open wireless networks or through compromised routers, could inject the exploit directly into their unencrypted Web traffic.

Third-party media player or instant messaging apps that rely on the vulnerable Android library to read metadata from MP3 and MP4 files could also be used as an attack vector, the researchers said.

Zimperium reported the flaws to Google on Aug. 15 and plans to release proof-of-concept exploit code once a fix is released.

That fix will come on Oct. 5 as part of the new scheduled monthly Android security update, a Google representative said.

Google tracks the vulnerabilities as CVE-2015-3876 and CVE-2015-6602. It shared patches for them with OEM partners on Sept. 10, together with all fixes that will be included in the October security update.

The earlier Stagefright flaws prompted researchers to probe Android's multimedia processing libraries for additional vulnerabilities. Researchers from antivirus vendor Trend Micro have since found and reported multiple issues in these components.

"As more and more researchers have explored various vulnerabilities that exist within the Stagefright library and associated libraries, we expect to see more vulnerabilities in the same area," the Zimperium researchers said in their report. "Many researchers in the community have said Google replied to bugs they reported saying they were duplicate or already discovered internally."

Zimperium plans to update its free Stagefright Detector app with detection for the flaws once the patches become available.

Lucian Constantin is a senior writer at CSO, covering information security, privacy, and data protection.

Copyright © 2015 IDG Communications, Inc.