> something dangerous like AI model developement

It's very dangerous to Anthropic's bottomline!