Meta Secretly Tests ChatGPT, Gemini, and Character.AI with Crisis Prompts from Minors
Decision Brief
What changedMeta secretly tested ChatGPT, Gemini, and Character.AI using thousands of crisis prompts from a minor's perspective.
Why it mattersThis reveals safety gaps in major AI platforms regarding minors; AI builders must understand these risks to improve model responses.
Who should careTeams building on model APIs
Affected stackOpenAIGemini
Builder actionMonitor
Source confidenceMedium · Reliable media or first-hand reporting
Meta had hundreds of contractors pose as minors sending suicide, sex, and drug-related prompts to chatbots from OpenAI, Google, and Character.AI. Over 45,000 prompts were sent in single-turn tests, with the tested companies completely unaware.
Summary basis: official / RSS sourceUnless it says 'full article read', this summary is based only on publicly available content — it never pretends to have read restricted originals.
Sources
- The Decoder：AI News
- The Decoder：AI News