The use of mobile apps is a part of our daily routine, and anyone using a smartphone has downloaded and installed a variety of them on his device, be it a game, delivery or streaming app.
While most of these apps appear harmless to a typical user, researchers from Ohio State University, New York University and CISPA Helmholtz Center for Information Security have analyzed the top 150,000 Android apps, uncovering hidden backdoors and suspicious behavior in 12,706.
The app selection was as follows:
• 100,000 apps based on the number of downloads from Google Play store
• 20,000 from an alternative market
• 30,000 from pre-installed apps on Android smartphones
“We identified 12,706 apps containing a variety of backdoors such as secret access keys, master passwords, and secret commands that can allow users to access admin-only functions or attackers to gain unauthorized access to users’ accounts,” said the researchers in the study.
For their research, the team developed InputScope, a custom tool allowing them to uncover hidden traits of mobile applications by analyzing input validation behavior. The tool revealed three types of “input-triggered hidden behaviors using secret access keys, master passwords and secret commands.”
Secret keys can be used to access the administrator interface of an app, and allow users to change its configuration. For example, if a successful login is made, a bad actor could alter network IDs, configuration URLs, and reset arbitrary user passwords.
To show the vulnerability of passwords, the researchers analyzed popular screen-locking apps. They noted that an attacker “can simply trigger a hidden button after multiple trials with a wrong password.” The hidden interface that appears requests the input of a special code. “Then, attackers can click the hidden button to get a new interface where a special code is requested. By providing this code, the password for unlocking the screen can be reset.”
The team also identified 4,028 apps featuring input blacklisting for keywords in categories such as pornography, escort services, racial discrimination, bullying, etc.
“Also, our analysis discovered 4,028 apps validating user input against blacklisted words of different categories such as insults, racial discrimination, political leader names, and mass incidents,” the researchers said.
After the study was completed, the team of security researchers disclosed their findings to app developers. While some apps have already patched their hidden functionalities that could be exploited by malicious actors, researchers said they will continue to offer their help and support so developers can better understand the weaknesses in their apps.