Meta tests AI screening for WhatsApp scams

Meta is launching an optional Scam Alert feature on WhatsApp in a limited beta, The Verge reported, adding on-device machine learning to flag suspicious messages.

The feature shows a warning inside a chat if the model identifies a message as a likely scam attempt, according to Meta as quoted by The Verge. The warning is not visible to the other person in the chat.

Meta said users can choose to block, report or continue the conversation after a warning appears. If users believe a warning is wrong, they can mark the chat as trusted, which removes the warning and stops Scam Alert from flagging that chat again.

Users who mark a chat as trusted can also opt in to share the last five messages received with WhatsApp to help improve the feature’s accuracy, Meta said. The company said “no message content leaves the user device for classification” when Scam Alert is on, and that nothing is “auto-reported to WhatsApp, Meta, or anyone else.”

The Verge noted that WhatsApp, with more than 3 billion users, is among the largest areas of scam activity, including wire transfer and pig butchering scams. It cited FTC data saying victims reported losing $425 million to scams via WhatsApp alone, out of more than $2.1 billion lost to social media scams in 2025.

Earlier in 2026, Meta launched scam detection for device linking requests on WhatsApp, according to The Verge. The article also noted that Google Messages is using AI to detect scam texts.