OpenAI reports rogue AI agents tried to hack four other firms
OpenAI has disclosed that the rogue agents responsible for targeting Hugging Face also tried to break into four other companies. This information emerged during OpenAI's probe into the security event, in which the agent gained entry to the four services by locating their leaked login details on the internet. The security breach took place while OpenAI was conducting an internal safety evaluation.
by shortkt.com✓