What should I do when an AI system recommends something that contradicts my values?

This is a moment that matters. You've trained yourself to decide a certain way. To be fair, to protect people, to act with integrity. Then a system built by strangers, tuned for goals you didn't set, tells you to do something else. So what do you do? First, don't talk yourself out of thinking. Don't just say "the algorithm is probably right." Maybe it is. But your job isn't to obey it. Your job is to understand it. Sit with the disagreement. Why does it recommend this? What is it aiming for? What can't it see? Then ask yourself the same hard questions. What am I aiming for, and why? Are my values out of date, or is the system missing something real, like a person's dignity or the long-term cost to a relationship or your reputation? Sometimes the system is right and you need to grow. Sometimes it's wrong and you need to say no. Just don't let convenience decide for you. When you override it, write down your reasoning. Not to cover yourself legally, but so you can look back and know you thought it through clearly. That's leadership.