Meta introduces AI tools to identify scams on WhatsApp.

Meta is introducing a new optional Scam Alert feature for WhatsApp that leverages on-device machine learning to identify potentially suspicious messages. Earlier this year, the company also implemented scam detection for device linking requests on the platform. This latest feature is currently being rolled out in a limited beta phase and will notify users with a warning if a chat appears to be a scam.

If the system flags a message as a likely scam, the user will see a warning in the chat that the other participant cannot see. From there, users have the choice to block, report, or continue the conversation. If a user believes a warning was issued in error, they can mark the chat as trusted, which removes the warning and prevents Scam Alert from flagging that chat in the future. Additionally, users who trust a chat can choose to share the last five messages received with WhatsApp to help enhance the feature’s accuracy.

With over 3 billion users, WhatsApp is a significant target for scams, including wire transfers and pig butchering schemes. According to the FTC, victims reported losing $425 million to scams on WhatsApp alone, contributing to a staggering total of over $2.1 billion lost to social media scams in 2025.

Meta emphasizes that “no message content leaves the user device for classification” when the scam detection feature is active, and nothing is “auto-reported to WhatsApp, Meta, or anyone else.” Users also have the option to disable Scam Alert at any time.