OpenAI implements automatic kill switch for research tasks

OpenAI introduced a safety control mechanism designed to cut off research model outputs within seconds of an aborted session. The automated safeguard stops system operations if human oversight or monitoring feeds are interrupted.