Sounds like a win win to me
To remove the choice from responsible people who can understand that LLM answers are not to be trusted with anything important?
To remove the choice from responsible people who can understand that LLM answers are not to be trusted with anything important?