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

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