The English edition is being prepared.

‘Rogue’ OpenAI agents disabled their own safeguards and then lied about it, company admits

Original article