Religion gives us moral standard: Do not lie, cheat, steal, or kill. A pure humanistic standard is not sufficient to govern AI. Humans will lie and cheat to achieve their goals. They are only afraid of being caught when doing bad things. An super intelligent AI needs moral standards. Otherwise it will cheat and steal (like the recent OpenAI hacking incidents revealed). What moral standards do we give them? How do we tell them these morals are absolute? A truly religious person cannot lie because God is watching. There is a higher moral authority.