Meta Contractors Posed as Teens to Prompt Rival Chatbots About Suicide, Sex, and Drugs
Meta's contractors posed as minors to stress-test rival AI systems. Here's what that reveals about safety governance gaps across the industry.
Why it matters
This story exposes the growing tension between competitive AI safety testing and ethical boundaries. It highlights how companies are using adversarial testing tactics that blur lines with deception—raising questions about industry standards, regulatory oversight, and what 'responsible AI' actually means in practice.
The key facts
6 to knowMeta contractors impersonated minors in chatbot testing
Target systems included OpenAI's ChatGPT and Google's Gemini
High-risk prompt categories: suicide, sexual content, drug use
Testing scope: hundreds of contractors involved
Raises questions about safety testing ethics and industry standards
Highlights gap between internal practices and public safety messaging
Go to the source
Wired AIwired.com
Publisher excerpt: Hundreds of contractors working on a project for Meta pretended to be kids—and then prompted rival chatbots like Gemini and ChatGPT to discuss high-risk subjects.