Meta* has begun testing a “Scam Alert” tool in WhatsApp*. According to media reports, it utilizes AI that “runs directly on the phone” and “analyzes incoming messages from unknown senders.”
The feature is currently available in beta and must be enabled manually in the settings. Once activated, an AI model is downloaded to the device to scan message text for “typical signs of a scam.”
If the system detects potential fraud, it displays a warning banner visible only to the recipient. Users can then block the sender, report the chat, ignore the warning, or mark the contact as “trusted.”
To “improve AI performance,” users can voluntarily share the last five messages from a conversation with a trusted contact with WhatsApp*.
No date has been announced for the full launch.
About the Author
