A number of tech companies agreed to develop and publish safety frameworks on how they will measure and combat risks associated with their AI models at the AI Seoul Summit.