The company has revealed details of AI model safety testing—including concerns about its new anthropomorphic interface.

Leave a Reply

Your email address will not be published. Required fields are marked *