Could an AI share state's secrets?


And if the answer is no, how can the AI dodge the question without any lie and making us doubtful?


Good question!

Perhaps to redirect back to an over all mission statement on working towards a peaceful future.

Maybe a really broad and vague answer about not taking sides and maybe something about establishing boundaries.