Replicant Devs Discover Backdoor in Samsung Android Devices
You may recall that about five months ago, we touched upon a study demonstrating how OEM modifications are the primary cause for most “Android” security issues. Unfortunately, we offer yet another example of OEM-caused security issues—but this time, it’s not because of an OEM skin or bloatware. Rather, this is a potential vulnerability at a far deeper level: proprietary modem software.
The OEM in question is none other than Samsung, the Android ecosystem’s largest and most successful device manufacturer, and the backdoor itself comes as proprietary radio software. This software is responsible for communicating with the modem hardware, and is capable of implementing RFS commands. These RFS commands are then able to perform I/O operations on the device’s storage.
No big deal, right? I’ll just load CyanogenMod and be done with it. Wrong.
Since the cause is a proprietary radio software, changing to an aftermarket ROM will not solve anything, so long as the ROM uses Samsung’s proprietary blobs. In fact, the Replicant team used Galaxy Note II and Galaxy S III devices running CM10.1 to demonstrate how this was ROM-agnostic.
Currently the list of known affected devices includes the Galaxy S, Galaxy S II, Galaxy Note, Galaxy Tab 2, Galaxy S III, and Galaxy Note II, but it’s highly likely that many other Samsung devices are vulnerable. Furthermore, this also seems to affect the Samsung-built Google Nexus S and Galaxy Nexus, as this is a back door at the radio software level, rather than as a part of an OEM skin. Whatsmore, on certain devices, this incriminated process runs as root.
While it is entirely possible that there is a legitimate reason for this backdoor, it’s hard to envision a scenario where one would be necessary. As such, it would be great to hear Samsung’s official statement on the matter. Until then, perhaps it would be a good idea to look into fully open source projects like Replicant, or at the very least, building an aftermarket kernel capable of blocking (and logging) RFS command requests.
You can learn more by heading over to the source link below.