Google Uses On-Device AI to Spot Scams in Chrome

Google is rolling out new artificial intelligence-based protections designed to detect online scams across its Chrome browser, Search engine, and Android operating system, the company announced Thursday. The initiative leverages Gemini Nano, Google’s on-device large language model, to enhance Safe Browsing features in Chrome version 137 on desktop platforms.

By processing data locally, Gemini Nano enables real-time analysis of potentially malicious websites, offering users immediate alerts about suspicious activity without transmitting data to remote servers. Google says this approach strengthens user privacy while improving the responsiveness of its threat detection systems.

The integration of AI-powered safeguards comes as part of the company’s broader efforts to combat increasingly sophisticated phishing schemes and online fraud. The rollout marks a shift toward decentralized security models, with on-device AI playing a central role in identifying threats.

The update is expected to gradually reach users as Chrome 137 becomes available, while similar protections are being deployed to Android and Search.