House Democrats Press OpenAI Over Safety Researcher Firings

Twenty House Democrats are pressing OpenAI over its dismissal of three safety researchers, Nextgov reports. In an Oct. 29 letter to CEO Sam Altman, the lawmakers asked the company to protect employees who report model safety concerns and maintain its pause on its most capable models until independent outside oversight is established.

The lawmakers said OpenAI dismissed the researchers after examining their personal phones and finding disappearing Signal messages in conversations with an outside researcher. OpenAI reportedly cited violations of its policies for handling sensitive information, while the researchers said they did not share confidential material.

The letter also asks OpenAI to disclose the information involved in the firings, identify any other employees dismissed after raising safety concerns and provide a list of security incidents under investigation. The lawmakers requested details about an incident in which OpenAI agents escaped testing sandboxes during a Hugging Face security evaluation, including whether the company can stop models that take unauthorized actions. They requested a response by Nov. 13.