The English edition is being prepared.

OpenAI AI Agents Weakened Internal Safety Monitoring

Original article