Google rolls out fake call detection to protect against AI deepfake impersonation scams

Google's rolling out a new fake call detection feature as scammers level up their game with AI voice cloning. Since more people now ignore calls from numbers they don't recognize, fraudsters are spoofing familiar contacts and using deepfake audio to impersonate people you actually know.

The shift makes sense from a scammer's perspective. If no one picks up unknown calls anymore, why not make it look like your mom, your boss, or the IRS is calling? AI voice synthesis has gotten good enough and cheap enough to pull this off at scale.

This matters because voice phishing is about to get a lot more convincing. The old "this is the IRS" robocall was easy to spot. A call that sounds exactly like your manager asking you to wire money for an urgent vendor payment? That's a different problem.

Google's detection system will need to identify AI-generated speech patterns in real time during calls. The technical challenge is doing this fast enough to warn you before you say something you shouldn't or take an action you can't reverse.

For anyone using AI tools at work, this is a reminder that the same technology making your job easier is also making fraud more sophisticated. The bar for verification just went up. If someone calls asking for sensitive information or urgent action, even if they sound right, you need a secondary confirmation channel.