What happened
Google is rolling out a new AI-powered security feature in Chrome designed to detect and block tech support scams directly on users’ devices. Built into Chrome 137 and powered by Google’s Gemini Nano large language model (LLM), the new tool scans web pages in real time to identify scam behavior like fake virus warnings and browser lockouts, tactics commonly used to trick users into calling fraudulent tech support numbers.
The system is part of Chrome’s ‘Enhanced Protection’ setting and will be enabled by default for users who opt in.
Going deeper
Tech support scams typically present alarming full-screen messages or pop-ups that falsely claim a user’s device is infected. Victims are urged to call a support number, which can lead to unnecessary charges or remote access that can result in theft or malware installation.
Chrome’s AI system uses Gemini Nano to analyze such scam signals locally, without sending browsing content to Google. If a scam is suspected, it sends anonymized LLM outputs and site metadata to Google’s Safe Browsing system for a second review. If confirmed, Chrome issues a warning to the user before any damage can occur.
Google states that this process is designed to preserve privacy and maintain performance. Resource management strategies include limiting the number of tokens used in each scan, asynchronous processing to avoid browser slowdowns, and GPU throttling.
The feature will arrive with Chrome 137, scheduled for release next week. It will also expand to Chrome for Android later in 2025.
What was said
“This is all done in a way that preserves performance and privacy,” Google said in the announcement.
The company clarified that Gemini Nano operates locally and only activates in relevant scenarios. Chrome’s design ensures minimal resource usage by controlling how often the AI runs and implementing safeguards against system strain.
Users can enable Enhanced Protection manually via Chrome Settings: Privacy and Security > Security > Enhanced Protection.
The big picture
Google’s move signals a shift toward AI-driven, real-time browser protection, especially on-device solutions that prioritize user privacy. As scams become more convincing and frequent, browsers are emerging as the first line of defense. If successful, Chrome’s new system could redefine how consumers are protected from online threats without compromising speed or data privacy.
FAQs
What types of scams will Chrome’s AI be able to detect in the future?
While the current focus is on tech support scams, Google hints at expanding detection to phishing sites, fake downloads, and other social engineering threats.
Will this AI feature work in Incognito mode?
Yes, if Enhanced Protection is enabled, the AI tool will function in both standard and Incognito browsing modes.
Can users customize or disable this AI tool?
The tool is tied to Enhanced Protection, which users can toggle on or off in Chrome’s security settings. There are no separate controls for just the AI feature.
Does this feature affect browser speed or battery life?
Google says the AI runs locally using lightweight processes, with resource safeguards like token limits and GPU throttling to minimize any performance impact.
Will enterprise or managed Chrome environments support this?
Google plans to include enterprise policy options for IT administrators, allowing organizations to control whether the AI scam blocker is enabled across managed devices.
Farah Amod
