MIT Department of Aeronautics and Astronautics

OpenAI and Anthropic model safeguards are flagging routine aerospace, robotics and security work as risky, costing developers time and forcing them to abandon chats or switch models, according to deve…