Microsoft is introducing a novel tool called the “scareware blocker,” which leverages machine learning and computer vision to tackle a widespread form of online fraud.
“Scareware” has plagued the internet almost since its beginnings. It’s frequently presented as fraudulent antivirus programs that falsely claim to have identified threats on a user’s device. These scams often manipulate users into installing harmful software or paying for unnecessary services.
Just last year, two tech support companies agreed to pay $26 million to settle charges from the U.S. Federal Trade Commission (FTC), which accused them of deceptive marketing practices, including the use of fraudulent Windows pop-ups to defraud customers.
“These organizations exploited fear and fabricated threats to deceive consumers, particularly the elderly, out of millions of dollars,” stated Samuel Levine, director of the FTC’s Bureau of Consumer Protection, at that time.
Microsoft already has mechanisms to block reported scam websites, but this new feature is specifically designed to detect and prevent previously unidentified scam tools from launching entire full-screen pages.
In November, Microsoft initially introduced the scareware blocker at its Ignite conference, and it is now encouraging users to participate in a preview program within the Edge browser to help test the functionality.
Setting Up
Users need to enable the scareware blocker through the “Privacy, search, and services” settings in Edge.
This feature adds another layer of protection to tools like Microsoft Defender SmartScreen, which already monitors suspicious activities on websites. The scareware blocker specifically addresses full-screen page attempts by scam tools, which can be particularly challenging to recognize and bypass; for instance, users might not realize they need to press the “Esc” key to exit full-screen mode.
Microsoft revealed that thousands of actual scam samples were utilized to train the machine learning model that powers the scareware blocker. This model employs computer vision to assess these samples against new scams as they arise in real time.
If the tool detects a potential scareware threat, it will revert the page from full-screen mode, halt any accompanying audio (such as alarms or alerts) associated with the scam, and prompt the user to either continue to the page or close it altogether.
Privacy
Concerns might arise regarding Microsoft’s use of computer vision to assess users’ screens. This draws some comparisons to Microsoft’s controversial AI-driven Recall feature, which captures screen images to compile a searchable history of user activity on their computers.
Nevertheless, Microsoft asserts that with the scareware blocker, the machine learning model operates locally on the user’s device, ensuring that no data is stored or transmitted to the cloud.
To enhance the model and the wider Defender SmartScreen application, Microsoft is actively seeking feedback from early users. Those using the feature will have the option to report scams by sharing screenshots with Microsoft, as well as flagging instances where the scareware blocker mistakenly tags a legitimate site as a threat.
Compiled by Techarena.au.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
