Once an Android user has turned AVS on in settings, it’s supposed to send information about any app being installed to Google for verification. Google ‘s servers then automatically respond with a result.
There are two levels of potential protection. If the service rates it a “potentially dangerous app,” it warns the user gets a warning that the app may harm the device if installed. If the service detects it as a “dangerous app,” it automatically blocks it from being installed.
But, as ZDNet reports, Google’s security service didn’t fare well in test performed by Xuxian Jiang, an associate professor of computer science at North Carolina State University.
The report says Jian installed 1,260 samples of malware on Google Nexus tablets, and Google’s service caught only 193 of them — a 15% detection rate.
But in a follow up by the New York Times today, Google notes that along with its verification service, it has a security system called Bouncer introduced earlier this year. Whenever an app is submitted to Google Play, the official Android app store, Bouncer puts it through a simulation on Google’s servers to search for hidden malware, spyware and trojans.
In a statement, a Google spokeswoman told the NYT that many of the apps in Jiang’s test were samples used by security researchers, and they are not downloaded by Android users. The company said its application verification service focuses on catching malware that people will actually encounter.
That said, Google clearly, like any other Internet-based service, sees it has to continually update its war chest in the battle against malware. For example, the search giant recently acquired VirusTest, a company that offers an online service for detecting malware.