Meta has been quietly building an AI-generated clickbait news feed inside its standalone Meta AI app, surfacing a stream of fully fabricated "stories" complete with AI-written text and AI-made images. The feature, a "For You" tab that's been live for at least several months, suggests article prompts that expand into puff-piece narratives when tapped. A Meta spokesperson told The Verge the company would pull the feature after being contacted for comment, though it's unclear how widely the rollback will be applied or how quickly.
The Meta AI app first launched in April 2025 with a public "Discover" feed showcasing AI-generated images and conversations from users, many of whom appeared unaware their outputs were being shared publicly. That feed has since been removed, replaced with a standard chatbot interface alongside the new For You page. According to Verge reporter Robert Hart, who was testing the feature from London, the algorithm aggressively served up British-flavored content: stories like "A royal butler finally settled the milk first debate" (verdict: tea first), "The psychology of joining a queue without knowing why," "The anatomy of the devastating British tut," and "Inside the extreme sport of visiting every UK pub." A colleague based elsewhere was served a different flavor entirely, with luxury watch bait like "My fake Rolex experiment" and "The brutal math behind the Rolex waitlist illusion."
The AI-generated text itself read like filler, restating the same premise over and over without ever building toward a real point, a hallmark of language model output stretched to article length. The approach mirrors a playbook Facebook has benefited from for years: low-effort, high-engagement clickbait tailored to whatever the algorithm thinks will keep a user scrolling. In this case, Meta has simply cut out the human middleman and is having the machines write the chum directly, then distributing it through a surface its users already trust. The episode is likely to fuel ongoing scrutiny of how AI tools are being used to generate engagement bait at scale, and how platforms decide what's worth removing once reporters start asking questions.