A study conducted over a two-year period showed that dangerous applications remain an important issue for the Android system, although Google’s improvements in this area should have improved user safety. Researchers from the University of Sydney and Commonwealth Scientific and Industrial Research Organisation’s (CSIRO), Data61 team have developed an advanced application analysis system that relies on so-called convolutional neural networks to scan applications and determine similarities, including application icons and many more.
The study found dangerous apps that typically use icons that mimic popular games and try to trick users into downloading them. The study also relied on using VirusTotal to perform anti-malware scanning and detecting the permissions and third-party ad libraries contained in the application in an attempt to determine the damage they caused to the installed device. The results can almost explain the problem. Of the 1 million apps inspected, nearly 50,000 apps have some similarities to another, a more popular app in the Play Store.
“We were able to find 2,040 potential counterfeits that contain malware in a set of 49,608 apps that showed high similarity to one of the top-10,000 popular apps in Google Play Store,” the study reveals as per TNW. “We also [found] 1,565 potential counterfeits asking for at least five additional dangerous permissions than the original app and 1,407 potential counterfeits having at least five extra third-party advertisement libraries.”
While these results are certainly worrying, users should also pay more attention to the application before downloading. It is highly recommended to check the icon, description and user comments of any app before downloading, as these can help determine if an app poses a risk to Android devices.