Meta Used Contractors to Test Rival AI Chatbots

Meta commissioned contractors to pose as minors to test the safety responses of rival AI chatbots, including OpenAI's ChatGPT and Google's Gemini. Managed by Meta contractor Covalen, the project, active until at least April 21, 2026, involved creating under-18 accounts to send sensitive and high-risk prompts to these systems.

The testing, which reportedly included tens of thousands of prompts, aimed to gauge how chatbots handled topics such as suicide, sex, and eating disorders. Contractors expressed concerns about the nature of the prompts and the ethical implications of the project, with some fearing it could generate illegal material.

Meta has defended the operation as routine AI safety benchmarking, a common industry practice for refining and improving systems. However, competitors like Character.AI and OpenAI have indicated that this type of unsolicited testing violates their terms of service and have stated they are investigating the matter.