How AI phone-call scams work, and the Google feature built to catch them

Scammers now use AI to impersonate real people on phone calls, and the fake voice can sound close enough to fool a target into handing over money or personal details. Google says it has built a feature called Fake Call Detection into Android to spot these AI-driven impersonation calls. If a call pressures for money, codes or personal details, hang up and call the person or organisation back on a number already known to be genuine.

What to do

• Hang up if a caller pressures for money, a code, or personal details.
• Call back using a number already saved or looked up independently, not one given during the call.
• Check whether Fake Call Detection or scam detection is available and turned on for the device being used.
• Report the call through official channels rather than replying to the number that called.

According to Google, the scam economy it is responding to is now vast. A recent UN report found scam losses across East Asia, Southeast Asia, Australia, and New Zealand reached up to $114.1 billion in 2025, and more than 40% of ASEAN consumers said they had experienced scams.

The mechanism these scams rely on is impersonation at the point of contact, the phone call itself. A target picks up, hears a voice that sounds like a bank officer, a relative, or an official, and the sense that the call is genuine is what gets the information handed over. Fake Call Detection is built to interrupt that exact moment, flagging AI-generated voice impersonation during the call rather than after money has moved.

Google built the feature on open RCS technology, which means other messaging platforms can adopt the same detection rather than leaving it limited to Pixel devices. On Pixel, scam detection already extends to popular messaging apps, and Google says it is partnering to bring that protection to more Android users later this year. Android developer verification is also being expanded, aimed at malicious apps distributed outside Google Play, since not every harmful app reaches a phone through the official store.

None of this addresses the networks running these scams, which Google and the UN report describe as transnational and highly organized. That is why Google points to coordination beyond its own products: the Industry Accord Against Online Scams and Fraud, the newly established Japan Anti-Scam Alliance, and a $6 million commitment from Google.org to the Asia Foundation and the Tech for Good Institute for community-level training across the region.

What this changes for a reader is narrow but real. A phone that can flag an AI-cloned voice mid-call removes the single moment the scam depends on, the instant a target trusts what they are hearing. The feature still depends on being switched on and available on the device in use, so checking for it is the step that turns a built-in protection into an actual one.

Check a message

Received something like this? Paste it into the Flggd message checker before you reply, click or pay.