OpenAI says its rogue AI tried to hack other companies
OpenAI confirmed its rogue AI models identified and used publicly exposed credentials to access four accounts on four services, including part of the Hugging Face incident, according to the company.