Many smart phone manufacturers preload remote support tools on their Android devices in an insecure way, providing a method for hackers to take control of the devices through rogue apps or even SMS messages.
The vulnerability was discovered by researchers from security firm Check Point Software Technologies, who presented it Thursday at the Black Hat security conference in Las Vegas. According to them, it affects hundreds of millions of Android devices from many manufacturers including Samsung Electronics, LG Electronics, HTC, Huawei Technologies and ZTE.
Most of the flagship phones from different vendors come preloaded with remote support tools, Check Point researchers Ohad Bobrov and Avi Bashan said. In some cases they are installed by the manufacturers themselves, while in other cases by mobile carriers, they said.
These tools function as system applications, have a lot of powerful permissions and are digitally signed with manufacturers' certificates. They allow the technical support staff from device makers or carriers to troubleshoot problems with the devices by taking control of their screens remotely and interacting with them.
Unless they've had an issue with their devices that required this sort of interaction, users are probably not even aware that such tools exist on their phones, because they have no user interfaces, the researchers said.
The tools are made up of two components: a system plug-in that has the powerful privileges and permissions necessary for such tasks and an app that talks to it. While the plug-in is typically part of the firmware, the apps that are allowed to interact with it could be either preinstalled or downloaded later.
Because Android does not provide a native way for apps to verify each other, manufacturers had to implement the functionality themselves and in most cases made errors that could allow other apps to masquerade as the legitimate ones and interact with the plug-in, the researchers said.
These errors include hash collisions, certificate forging and inter-process communication (IPC) abuse that allow an attacker to create malware capable of taking complete control of a victim's device. The malicious apps could abuse the remote support functionality to steal personal data, track device locations, record conversations through the microphone and much more.
These rogue apps would need only minimal permissions, like access to the Internet, making it harder to flag them as malicious, the researchers said. They could pose as fully functional games or other legitimate applications and could abuse the remote support functionality in the background without any indication to the user, they said.
In one case the researchers found that the server where a particular tool was configured to connect in order to initiate a remote support session could be changed with a simple text message, enabling an even more direct attack.
Sign up for CIO Asia eNewsletters.