Meta paid contractors to pose as teens and probe rival AI chatbots

What happened

Meta hired contractors to create fake teen profiles and interact with rival AI chatbots such as OpenAI’s ChatGPT. The project, known internally as Cannes and managed by a contractor called Covalen, had hundreds of workers posing as users under 18. They sent messages, prompts, and images to competitor bots, then recorded the responses in spreadsheets for analysis. This spying effort continued until at least April 2026.

Why it matters

Meta’s tactic exposes the high stakes and cutthroat competition in AI chatbot development. By impersonating underage users, Meta aimed to probe rivals for weaknesses or gain insights into their models’ behavior with a vulnerable demographic. This raises ethical questions about user privacy and the lengths companies will go to stay ahead. For builders and operators, it underscores the risk of adversarial probing that might exploit chatbot weaknesses or feed data back to competitors.

The approach also highlights the pressure to anticipate and counter competitors’ growth in conversational AI. Meta’s investment in contractors to systematically test rivals indicates that simply releasing a strong model is not enough. Monitoring and stress-testing opponents online has become a tactical business imperative, but it may also prompt tighter regulations or scrutiny over data practices and how chatbots are evaluated.

What to watch next

Watch for how Meta and other major AI players adjust their surveillance, testing, and competitive intelligence strategies going forward. The public disclosure may spark legal or regulatory investigations around user impersonation and data collection methods. It could also trigger faster development of tools to detect and block coordinated probing attempts on AI platforms.

For chatbot operators, expect rising demand for security features that differentiate genuine users from contractors or bad actors influencing model training data. Investors should track whether this pressure accelerates consolidation or forces startups to fortify their tech and policies to survive in a surveillance-heavy AI market.

AI Quick Briefs Editorial Desk