The developers from Microsoft added an instrument called Scareware Blocker to their company browser. He involves machine learning and computer vision to combat the so-called strap programs that encourage users to load malicious software under the pretext of fighting viruses.
Image source: Microsoft
Strashka programs have existed on the Internet for many years. Most often they are given out for fake antiviruses that appear in a pop -up window and report that a threat to security was supposedly discovered on the user’s computer. The creators of this are trying to encourage users under the pretext of removing the virus to download their software, which is actually viral, or pay for the useless application.
Microsoft has already added a number of functions to EDGE aimed at blocking fraudulent sites identified earlier. However, the new security function is expected to block even previously unknown fraudulent resources when they try to open a full -screen message on the user’s computer. For the first time, Microsoft spoke about this AI function at the IGNITE conference in November last year. Now the software giant has launched testing this tool among the participants in the pre -evaluation of Edge.
The new function will add an additional level of security to what already has in Microsoft Defender Smartscreen, which can recognize suspicious activity on web pages. The pages of pages-straps will work when the fraudulent application tries to expand full-screen message. Microsoft said that thousands of real cases of fraud were used to teach the algorithm. In addition, this function involves computer vision to compare already known cases of fraud with a new suspicious activity that will be exposed online.
The fact that Microsoft uses computer vision to analyze what is happening on user screens can cause concerns about confidentiality. Microsoft claims that the algorithm operates locally on the user device and does not transmit data to remote servers.