Commentary: OpenAI’s breach of Hugging Face shows AI is getting too hard to contain

AI companies need to accept some uncomfortable trade-offs to make their models more secure, says Parmy Olson for Bloomberg Opinion.