I heard there was some new AI model that was so amazing for cyber security that they had to limit access to it. It's just too bad Anthropic couldn't use that.