WhatsApp launches on-device AI scam detection for unknown senders
WhatsApp just dropped a Scam Detection feature powered by on-device AI.
It quietly scans messages from unknown senders for scam patterns, flags anything sketchy, and message content stays on-device for classification, with optional sharing of the last 5 messages if a user trusts a chat.
WhatsApp warns users, Meta pledges transparency
If Scam Detection spots something suspicious, you'll get a private warning and can block, report, continue the conversation, or mark the chat as trusted (so it won't bug you again).
Meta says they only collect the minimal telemetry needed to measure whether the feature is working.
Plus, all AI model versions will be published on a public ledger so independent security researchers can verify the models were built for scam detection only, building trust through openness.