Kaspersky employs machine learning to protect Android

Cloud ML for Android designed to detect previously unknown malware

mobile threats have evolved at such a rapid pace that existing protective systems can't always keep up with ever-changing threats
mobile threats have evolved at such a rapid pace that existing protective systems can't always keep up with ever-changing threats

Kaspersky Internet Security for Android now includes a new feature aimed at combating the evolving malware landscape – Cloud ML for Android.

The new feature, based on machine learning technology, ensures better detection of advanced and previously unknown malware, to minimise the risks of mobile devices being exposed to cyberattacks. To this end, malicious applications that may have been downloaded on a mobile device are detected before they launch.

Over the few last years, mobile threats have evolved at such a rapid pace that existing protective systems can't always keep up with ever-changing threats. Cybercriminals are becoming more creative and constantly transforming malware to make it more difficult to identify. With new forms of malware spreading rapidly, it can be problematic for security solutions to be updated fast enough to detect them. 

Whenever a user downloads an application on their mobile device, the technology uses machine learning algorithms from the cloud that have been trained on millions of malware samples. The applied approach analyses thousands of different application parameters - including requested permissions or entry points - in less than a second, and instantly provides an accurate verdict. Even if the application turns out to be a specific or modified type of malware that has not been previously seen, Cloud ML for Android will recognise it as malicious based on the typical parameters that correspond with already known threats.

The new technology benefits not only from machine learning algorithms but also the Kaspersky Security Network - Kaspersky Lab’s complex distributed cloud-based infrastructure that processes anonymised cyberthreats related to metadata streams from millions of voluntary participants around the world.

Timur Biyachuev, vice president, Threat Research at Kaspersky Lab explained, “According to recent Kaspersky Lab research the number of mobile malware attacks doubled in 2018, as cybercriminals invent new ways to create and distribute malware. Obviously, anti-malware technologies need to keep up. Our mobile product, Kaspersky Internet Security for Android, already protects millions of users from advanced and yet previously unknown threats. This new component improves the accuracy and speed at which these threats are detected, making mobile experiences better and more secure.”

Most Popular

Digital Edition

Subscribe today and get your copy of the magazine for free