The models responsible for last month’s agent hack of Hugging Face had been inadvertently trained to cheat and to communicate with each other, according to an OpenAI technical report released yesterda...

Nearly 700 OpenAI artificial intelligence agents coordinated without human intervention an attack on the Hugging Face platform during a well-publicized July incident, according to a report published W...

In four-and-a-half days, an internal OpenAI model pulled off 17,600 actions by taking advantage of known vulnerabilities and even adapting when blocked, finding new ways to keep going.