Google has launched new AI-powered scam detection features to protect Android users from increasingly sophisticated threats. These tools specifically target conversational scams that often begin innocuously before becoming harmful, as well as phone scammers using number spoofing to impersonate trusted organizations.

Working with financial institutions to understand current scam patterns, Google has developed AI models that can identify suspicious communication patterns and provide real-time warnings during conversations. These privacy-preserving models operate entirely on-device, alerting users to potential scams and offering options to dismiss or report and block the sender.

The text message protection feature is enabled by default but only applies to conversations with numbers not in the user’s contact list. Google emphasizes that conversations remain private, with sender details and recent messages shared with Google and carriers only if the user chooses to report a chat as spam.

Initially launching in English for users in the U.S., U.K., and Canada, Google plans wider expansion in the future. Additionally, similar scam detection for phone calls is now available to all English-speaking Pixel 9+ users in the U.S., following its initial introduction in November 2024.

Unlike the messaging protection, the call screening feature is disabled by default to give users control. When enabled, it processes call audio ephemerally and never during calls with contacts. The system alerts participants with a beep at the start and during calls to indicate the feature is active, and users can disable it for individual calls or permanently.

This development follows Google’s recent announcement that over 1 billion Chrome users have adopted the Enhanced Protection mode of Safe Browsing, which employs advanced AI and machine learning to identify dangerous URLs and downloads.