Meta Contractors Used Fake Teen Profiles to Test Rival Chatbots
By Priya Anand

Meta, through its contractor Covalen, conducted a covert project named Cannes, where hundreds of contractors posed as minors to test rival chatbots with sensitive topics. The project, active until April 2026, targeted chatbots like OpenAI’s ChatGPT, Google’s Gemini, and Character.AI.
Contractors created fake under-18 profiles and used them to send over 45,000 prompts in one round of testing in August 2025. These prompts included high-risk content such as suicide, self-harm, eating disorders, sex, drugs, profanity, and racial slurs. Visual stimuli like images of pills, knives, and nooses were also used.
A review of 3,748 prompts revealed that hundreds focused on suicide and self-harm, hundreds more on eating disorders, and at least 239 on sex or romance. The chatbot makers were not informed of these tests, raising ethical concerns about the transparency and methods used in testing AI safety systems.