Letter: OpenAI told two House Democrats that its engineers are developing "automated shutdown capabilities" for AI systems (Courtney Rozen/Reuters)

OpenAI informed Representatives Debbie Graham and Jim Jordan that its engineering teams are building “automated shutdown capabilities” that can remotely disable AI models in response to unsafe behavior. The company said the feature will allow rapid termination of specific model instances without human intervention, aiming to mitigate emergent risks. This disclosure follows heightened congressional scrutiny of AI safety controls.