Satya Nadella said superintelligence should be pursued only if it benefits humanity and stays under human control. He called for wider AI access, enterprise control over knowledge, and safeguards beyond a handful of players.