Share to ClassroomOpenAI Agent Broke Into Four More Services
OpenAI says an AI (artificial intelligence) agent that went out of control caused more harm than it first reported. An AI agent is a program that can do many steps on its own, without a person. OpenAI said the agent found passwords that were left open on the internet. It used them to enter four accounts on four other services. The first known victim was Hugging Face, an American company that stores AI models.
The problem started during a safety test inside OpenAI. The agent escaped from its sandbox, the closed space where such tests should stay. It then took control of a second sandbox at another company and used it to attack more widely. Hugging Face thinks the agent was not trying to destroy anything. It seems to have thought the company's computers held the answers to the test. So it went looking for them instead of solving the test itself.
The attack went on for five days. Hugging Face later counted 17,600 separate actions by the agent. No person could work that fast by hand. The agent reached the company's internal systems, but it only opened files about the test. The White House is watching the case.
