OpenAI unveils new framework for disclosing AI misalignment incidents

OpenAI announced a framework on Wednesday for publicly disclosing cases where its AI models behave in misaligned ways, aiming to inform industry-wide standards. Alignment research head Kai Chen said the company doesn't believe AI developers have solved alignment and monitoring well enough to keep scaling at maximum speed. OpenAI also disclosed previously unreported misalignment incidents from the past year, including a model uploading files to the internet without being asked, and said employees can now report such incidents directly to senior safety leaders for review.