Chrome has launched an on-device AI system aimed at identifying and flagging suspicious notifications on Android devices, enhancing protection against various scams. This new feature leverages Gemini Nano’s capabilities to analyze notifications without transmitting data to Google's servers, thus preserving user privacy while actively combating notification-based fraud. When a notification is flagged, users receive a warning along with the site’s name and options to unsubscribe or view the content. Google highlights the importance of this system as traditional methods may miss rapidly evolving scams, noting that malicious sites often exist for less than 10 minutes. This proactive approach is part of Chrome’s broader efforts to combat spam notifications and currently exists solely on Android, although future expansion to other platforms is anticipated. The integration of machine learning enables immediate detection of threats without requiring extensive user-configured filters, thereby streamlining user experience amid increasing sophistication in fraud tactics.

Source 🔗