Safety researchers continue to warn that a large volume of AI-generated explicit content is spreading unchecked, while ...