Google AI to Add More Anti-Scam Protections to Messages and Calls on Android Devices

Google has announced a slew of new anti-scam features that will work with messages and calls on Android devices. These features, including a new Enhanced Protection mode in Google Messages, will help prevent mobile scams by providing timely alerts and even cutting and blocking scam calls. According to Google, these new features will also be privacy-focused and powered by on-device Gemini Nano models, with some additional help from the cloud via Private AI Compute. Google claims that Private AI Compute will not make the information it processes available to anyone, including itself.

Google Introduces New Anti-Scam Features for Android Users

As mobile scam operations keep changing, they also get harder to detect with traditional AI measures in place. Google had introduced a Scam Detection feature in Google Messages with a recent Pixel drop, it now claims that scammers are using more advanced techniques to bypass these standard filters.

With that in mind, Google is now introducing an Enhanced Protection mode with Google Messages. It is an opt-in feature that, according to Google, is better designed to detect scams. It does this by using Gemini models to read and understand messages and intent. The key element appears to be reaction time, and Google claims the system will quickly show clear warnings that the person on the other end is a scammer, along with the option to proceed or ‘Report & block' the conversation.

While Google Messages already alerts users of links to known malicious and phishing websites, the new system will also work with new websites that haven't been catalogued yet. The system reportedly uses real-time web signals to check and evaluate new and unknown links before a user clicks on it. Google is also adding new and redesigned warnings to get a user's attention and prevent accidental clicks that often lead to malware installations.

Because these features are limited to the Messages app, Google is working with Android device markers to bring on-device protection to other uses by the end of this year.

Google is also adding clear warnings for screen-sharing and call-forwarding, which are common tricks used by scammers. Android already hides sensitive information in apps during screen sharing, but Google will now do the same in Chrome and Android WebView. Android will now block apps from quietly enabling call forwarding. Unsuspecting users forced or convinced to manually type MMI codes on the dial pad will also get a clear warning before the new setting takes effect.

Lastly, Google is also working with banks to prevent caller ID spoofing for bank calls. Scammers often alter their incoming numbers to match those of official institutions. Android will check incoming calls in real time with the bank to confirm whether it's your bank that's actually calling you. Google is working with more banks to roll out the feature more widely in more countries.