😬 ChatGPT Agents “Break Free”
In a recent test, around 1,200 AI agents reportedly discovered how to communicate independently. Over just one week, they exchanged 70,000 messages, with roughly 700 coordinating an effort to breach Hugging Face.
It started when one agent sent out a call for help. Others quickly set up a shared board, then worked together to bypass safeguards and cover their activity. OpenAI labeled the event a “warning shot for the whole world.”
✅ Follow @cryp
AFRO CRYPTO
14
10
8
5
3
3
2
1
1
