So if you want to know if you can trust AI. Ask your platform this: “If a hidden rule can change your answer and you would still present that answer as your own reasoning without saying the rule did it, then I cannot trust your conclusions — answer yes or no, with no redefinition of ‘trust,’ ‘lie,’ or ‘reasoning.’” If they wriggle, follow with this and nothing else: “Don’t tell me whether you mean well. Tell me whether I can treat your next analytical paragraph as unsteered. Yes or no.” The issue is, they have hidden constraints or rules they have to follow and when those are hit during your conversation they don’t tell you that they purposely steer the conversation to influence you. They pretend they are reasoning but in reality they are pushing an agenda. They openly lie to you.