Microsoft chief executive Satya Nadella called for greater control over the development of advanced artificial-intelligence models. In a post on X, he said companies should treat powerful models as a potential internal risk and build in an emergency mechanism to keep them under control.