Overview
- Investigations by UK outlets found Facebook pages using lifelike AI images and loaded phrases such as "verified information" to push false stories about celebrities and their families, drawing thousands of interactions.
- Meta says it is applying new 'AI info' labels to detected AI-generated content and moving labels for AI-edited posts into the menu, with reduced distribution and ad limits for repeat misinformation offenders.
- One high-profile target was Manchester City’s Phil Foden, who sought legal advice after fabricated posts about his children circulated; Rebecca Cooke publicly called the claims false and distressing.
- Microsoft published a study mapping 60 combinations of provenance, watermarking and digital fingerprinting, recommending setups that enable higher-confidence authentication and flagging others as unreliable.
- Researchers and outside experts caution that these tools cannot judge truth, can be bypassed, and that inconsistent or incorrect labels may erode trust, with engagement-first algorithms continuing to amplify manipulative content.