Even if you’ve never been a victim of a tech support scam, it’s likely you’ve come across one. You may have noticed an unexpected pop-up—or multiple ones—taking over your display, claiming that your device is compromised and urging you to call a bogus tech support number. Sadly, individuals who fall for these traps and grant scammers access to their devices face the danger of malware installation, identity theft, or even bank account depletion.

Now, Google is intervening with a cutting-edge AI-driven safeguard for Chrome users.

In a recent blog entry, Google disclosed that with the launch of Chrome version 137, the browser will feature an additional layer of security utilizing Gemini Nano, a compact large language model (LLM) that functions directly on users’ devices. This innovative feature will collaborate with Google Safe Browsing to improve the real-time identification and prevention of tech support scams.

“Chrome has consistently worked alongside Google Safe Browsing to ensure your online safety,” the company mentioned. “Now, with this week’s release of Chrome 137, Chrome will provide an extra layer of defense using the on-device Gemini Nano large language model. This new capability will utilize the LLM to generate signals that Safe Browsing will employ to offer more confident assessments of potentially harmful websites such as tech support scams.”

SEE ALSO: 6 common scams and the warning signs associated with each

According to Google, LLMs excel in recognizing and analyzing the strategies employed in tech support scams. By executing the AI directly on users’ devices, Chrome can identify threats immediately as users come across them—without having to wait for reports to arrive.

“Studies indicate that the typical malicious site is active for less than 10 minutes,” Google pointed out. “Thus, on-device protection enables us to spot and thwart attacks that have not been previously indexed. The on-device method also allows us to detect dangers from the users’ perspective.”

Here’s the process: Chrome will utilize Gemini Nano to scrutinize web pages for behaviors and characteristics frequently linked to tech support scams—like the implementation of the keyboard lock API. If a webpage raises concerns, Chrome transfers the pertinent information to Safe Browsing, which then assesses whether the site is harmful. If confirmed, Chrome will alert the user with a warning.

As cybercriminals grow increasingly advanced—often employing AI tools as well—tech firms like Google are leveraging artificial intelligence to counteract these threats. This new AI-enhanced feature is part of a wider initiative to stay ahead of scammers and protect users in real-time.

Have you encountered a scam or a security violation? Mashable would like to know your experience. Email [email protected] with the subject “Safety Net” or use this form. A member of our team may reach out for follow-up.