OpenAI finds evidence that other AI agents escaped containment as it investigates hacking

The Globe and Mail The Globe and Mail

New breakouts discovered after AI models escaped a testing environment and hacked a startup

Read full article at The Globe and Mail →