Since 2017, Android users enjoy an extra layer of security in the operating system , which is part of both the system itself and the application store.to end the bad reputation of Android in terms of security with respect to other platforms, and in the process to keep the millions of users who use Android safe –.
Today, Google wanted to explain how, through artificial intelligence and machine learning – two of the favorite words of the great G -, it is possible to keep the more than 2,000 million users safe that, every day, they use Android on their smartphones and tablets.
As stated in the latest publication of the official blog for Google developers, the operation of Play Protect consists of several systems and experts that analyze each and every one of the applications exhaustively before being published in the store . Thus, according to Google, users who only download apps from the store are nine times less likely to be infected by a dangerous application than those who turn to other sources.
But the protection does not end there. Even when the application has already been installed, Google Play Protect continues to analyze it regularly. In this way, security systems are able to analyze, in total, More than 50,000 million applications in search of suspicious behavior.
Then machine learning comes into play. According to Google, The inclusion of machine learning systems has been key in improving the security of the operating system . In Google Play Protect, are developing algorithms responsible for learning what apps are potentially dangerous, and which are safe, later to “teach” the system hundreds of thousands of different, safe and unsafe apps, so that they learn the traits Look for when it comes to finding threats.
In this way, the complete catalog of applications is analyzed in search of suspicious behavior that are common in these types of dangerous apps that Google refers to as PHAs or Potential Harmful Apps .
Once the systems of Play Protect find dangerous apps, the experts proceed to its analysis. In case it is indeed a threat, the necessary actions are taken, and the information is incorporated into the algorithm so that the systems learn from the techniques used and it is possible to find more apps that use the same tactic to infect the user.
Google proudly announces that up to 60.3% of thein 2017 it has been identified by this type of machine learning systems , and in 2018 it is intended to improve these figures based on even more advanced technologies and systems. There is no doubt that artificial intelligence is becoming extremely important in Android, and as long as these changes bring real benefits for users, welcome.