In a move against increasingly sophisticated fraud, Meta announced a new set of AI-powered protections for WhatsApp, Facebook, and Messenger. The initiative responds to a sharp rise in scams using AI-generated profiles and deceptive websites. The company reported removing over 159 million scam ads last year alone.

The rollout begins on WhatsApp, a frequent target due to its device-linking feature. Scammers often trick users into sharing linking codes by posing as relatives or service providers. Once inside an account, they impersonate the user. Meta terminated nearly 7 million such accounts in 2025. The new system will now trigger additional warnings when an unlinked device in an unfamiliar country attempts to connect, prompting users to reconsider.

On Facebook and Messenger, similar protections will flag suspicious accounts for review. Warnings may note if a profile is very new or located in an unexpected country. Messenger will specifically monitor for patterns in job-offer scams designed to steal data.

To identify fraudulent material, Meta is applying AI scanners that review content for hallmarks of deception, including fake celebrity images and phishing links. The company states these tools analyze activity patterns to spot fraud but has not detailed its data handling processes. This lack of specific disclosure revives longstanding questions about privacy and oversight when monitoring expands on personal messaging platforms. Meta has not provided a public timeline for a full global release.

Source: CNET