TechReport: Researchers breached OpenAI with Claude

A report alleges an intrusion into OpenAI in which researchers leveraged Anthropic’s Claude; they entered a staff member’s ChatGPT account and inspected files holding company secrets.