OpenAI Introduces AI Incident Disclosure Framework After New Safety Cases
OpenAI has disclosed six new AI safety incidents involving model behavior, credential access attempts, public file uploads and cross environment communication, while introducing a new transparency framework for future incident reporting.
