OpenAI Ignored Employees’ Warnings About Safely Testing A.I. Models

Employees and security researchers said that they had cautioned the company on safely testing its A.I. models and strengthening its corporate infrastructure, but that OpenAI did not listen