Pixel phones now have AI-powered scam protection

Google's phones can monitor for scam calls in real time, entirely using on-device hardware to recognize malicious patterns.

When’s the last time you got a call from an actual, genuine person that you don’t already know? I’m struggling to remember anything in the last three months that wasn’t a scammer, either some poor schmuck working in an organized crime call center or just a recording on the other end of an autodialer. Phone makers and service providers aren’t unaware of this, and Google’s latest attempt to fight back is just for Pixel owners.

According to the official Google Security Blog, the company is now bending the might of it Gemini AI tools towards protecting Pixel devices from scammers in real time. “Scam Detection uses powerful on-device AI to notify you of a potential scam call happening in real-time by detecting conversation patterns commonly associated with scams,” says the news post. The system is in beta, rolling out to Pixel 6 and newer phones starting now.

I assume the phrase “we’ve been trying to contact you about your car’s extended warranty” makes about a million digital alarm bells go off, but Google specifically mentions an attempt to phish users with a phony bank account breach.

If the on-phone system detects a scam, you’ll get an audio chime, a visual notification, and a haptic vibration — the modern phone trifecta approximating a Red Alert on the Enterprise. Maybe it seems like overkill, but considering the frequency of these scams, I don’t think you can be too careful. I happen to know a computer security professional with decades of experience who fell prey to one of these in a moment of inattention, to the tune of several thousand stolen dollars. Nobody is immune.

Google says that the new system uses Gemini Nano, running entirely on local Pixel hardware. That’s one of the reasons it’s exclusive to the Pixel 6 and newer models. (A suspicious type might also wonder if this stuff could run on any Android device with an NPU…but I digress.) The recordings of conversations never leave your phone’s isolated hardware, which is comforting, but still requires a degree of trust.

If you don’t think Google has earned that trust (and who could blame you?), you’ll be even more comforted by the fact that this system is turned off by default.

Google is also expanding its protection system against malicious Android apps, with a similar always-on detection system added to Play Protect. The post says that it can add real-time alerts based on the “activity patterns” of apps, making note of newer malware techniques that let the app lay dormant for a while before springing into action. This feature is also starting out on the Pixel 6 and newer, but Google says it’s coming to “additional phone makers in the coming months.”