OpenAI cuts ties with 3 researchers over alleged misconduct

OpenAI has terminated three safety-team researchers for allegedly disclosing confidential company information to an outside monitor, according to The Wall Street Journal. The company accused the staffers of violating policy and "breaking the trust essential to our work." The ousters come amid growing tension over how AI companies balance intellectual property protection with the need for independent safety audits, and are reported to coincide with the shelving of the GPT-6.1 Astra model after internal safety and alignment failures. Commentators note that broad NDAs can deter employees from reporting serious safety risks to regulators or independent evaluators, prompting calls for ethical NDAs and clearer whistleblower protections.

(Source：LinkedIn)