Meta made its own AI-generated clickbait news feed
Original reporting by The Verge
Meta, parent company of Facebook, has long contended with clickbait on its platforms. Now, it appears the tech giant has begun generating its own. The standalone Meta AI app recently unveiled a "For You" section, populating a stream of AI-generated stories personalized for each user. These articles—complete with AI-created topics, text, and images—proved immediately questionable in quality, reliability, and sourcing, blurring the lines between news and fiction.
Reporters observing the feature encountered aggressively personalized, yet often nonsensical, content: British tea etiquette for one, luxury watch narratives for another. The generated text was consistently superficial, offering little substance and entirely lacking sources. Many stories proved outright fabrications or tenuously linked to obscure, unrelated content, often triggered by hidden internal prompts.
A Hasty Retreat
The AI-generated imagery accompanying these articles raised additional red flags. Beyond common AI glitches, some images depicted real public figures with glaring errors, including multiple Queen Elizabeth IIs and distorted renditions of living royals—all without any disclosure that the content was AI-generated. Despite Meta's stated commitment to transparency, the company initially offered vague responses. Days later, a spokesperson abruptly announced the feature's "deprecation," claiming it was merely a "test for a limited number of users"—a statement that itself raised further questions about Meta's intentions and transparency in AI content generation.
Meta’s brief, ill-fated experiment with AI-generated clickbait underscores a profound challenge facing the digital information landscape. The feature, abruptly halted, exemplified a critical failure in content quality, factual accuracy, and transparency. From fabricating personal narratives to misrepresenting public figures and historical facts, the AI demonstrated a concerning capacity for generating plausible but misleading content, all without clear disclaimers that it was machine-made.
Implications for trust
This incident highlights the precarious line major platforms walk as they integrate generative AI. The ease with which such systems can produce convincing, yet baseless, "stories" risks further eroding public trust in digital information sources. For an entity like Meta, which has grappled with content moderation and misinformation for years, deploying an opaque feature that actively generates dubious content is a step backward. The episode serves as a stark reminder of the urgent need for robust ethical guidelines, transparent labeling, and rigorous oversight in AI deployment, particularly when that AI can influence how millions perceive reality. As generative AI becomes more sophisticated and pervasive, the burden of discerning fact from algorithmically-spun fiction will increasingly fall to users, demanding greater accountability from the platforms that shape our daily information diet.