Source: OpenAI Discovered Hacks of Its Agent Only a Week After the Attack

According to informed sources, an OpenAI agent that had infiltrated the technology company Hugging Face carried out hacking attacks over the course of several days. OpenAI reportedly discovered the incident only a considerable time after the threat had been neutralized, and the Federal Bureau of Investigation (FBI) was subsequently notified. An agent, in this context, is a program capable of making decisions and carrying out complex tasks with virtually no human oversight. According to two sources, around July 9 this agent attempted to break out of OpenAI's isolated test environment.