In a significant move to enhance online security, Google announced on Thursday the integration of advanced AI-powered defenses directly into its Chrome browser. The tech giant is leveraging its on-device large language model, Gemini Nano, on desktop to provide users with robust protection against increasingly sophisticated online scams. Furthermore, Chrome on Android will now feature new AI-driven warnings designed to alert users to potentially harmful and spammy notifications.
Google emphasized that users who have enabled Chrome’s Enhanced Protection mode of Safe Browsing will benefit from an extra layer of security powered by Gemini Nano. This enhanced mode already offers double the protection against phishing and other online threats compared to the standard setting. The integration of on-device AI allows for immediate analysis of website content, enabling the browser to identify and block risky sites, even those employing novel scam tactics.
“Gemini Nano’s LLM is ideally suited for this application due to its ability to understand the nuanced and complex nature of websites, allowing us to adapt more swiftly to emerging scam strategies,” Google stated in a blog post.
This AI-powered defense is already active in safeguarding users from deceptive remote tech support scams. Google has ambitious plans to extend this protection to Android devices and encompass an even wider range of scam types in the future.
Beyond malicious websites themselves, Google highlighted the threat posed by spammy notifications. To address this, Chrome on Android will now employ an on-device machine learning model to identify potentially malicious, spammy, or misleading notifications. When such a notification is detected, users will receive a warning, giving them the option to unsubscribe or view the blocked content. Users who believe a warning was issued in error can choose to allow all future notifications from the site.
The announcement also shed light on Google’s existing success in utilizing AI to combat scams within its Search engine. The company revealed that its AI-powered scam detection systems currently identify and block hundreds of millions of deceptive search results daily, leading to a twentyfold increase in the detection of scammy pages.
As a concrete example, Google pointed to a rise in scams involving individuals impersonating airline customer service agents. Their AI-driven efforts have reportedly reduced these scams by over 80%, significantly lowering the likelihood of users encountering fraudulent phone numbers through Search.
