Meta isn't just removing bad ads anymore — it's trying to stop you from getting fooled in the first place.
The company announced a wave of new anti-scam tools rolling out across Facebook, Messenger, and WhatsApp, marking a shift from purely reactive enforcement toward proactive, real-time user warnings.
The timing isn't coincidental: in 2025 alone, Meta pulled more than 159 million scam ads from its platforms, 92% of which were caught before anyone reported them.
The centerpiece of the new rollout is AI-powered impersonation detection. Rather than waiting for users to flag fraudulent celebrity endorsements or fake brand pages, Meta's systems now analyze behavioral and visual signals to identify "celeb-bait" content — those deceptive posts that slap a public figure's face on a crypto scheme or miracle supplement. Traditional keyword filters routinely miss these; the new AI approach targets the deceptive framing itself.
On Facebook, users will start seeing friction warnings when accepting friend requests that show suspicious signals: accounts that joined recently, appear to be in a different country, or share few mutual connections. The interface surfaces these red flags directly on the confirmation screen, along with options to report or block without penalty.
WhatsApp gets arguably the most technically interesting addition: a device-linking alert. Scammers have been socially engineering users into linking their WhatsApp accounts to attacker-controlled devices — essentially hijacking the account without cracking a password. WhatsApp will now flag linking requests that trigger behavioral anomalies.
Messenger's scam detection, previously limited in geographic scope, is also expanding internationally, with on-device AI (machine learning that runs locally, not on Meta's servers) flagging suspicious conversation patterns and prompting users to request a human review.
On the advertiser side, Meta is pushing toward a world where verified advertisers account for 90% of ad revenue by the end of 2026, up from current levels, tightening the pipeline for fraudulent ad buyers in high-risk categories.
Meta also disclosed a joint disruption operation conducted with the FBI, the DOJ Scam Center Strike Force, and the Royal Thai Police, alongside takedowns tied to romance scam networks and a scam center in Nigeria dismantled with the UK's National Crime Agency.
For ordinary users, the practical takeaway is simple: the warnings are there now — don't dismiss them.