Satya Nadella called for advanced AI systems to include containment and an emergency brake. He said humans should be able to pause or shut a model down mid-task. He also demanded independent controls and industry standards. He wrote that frontier models should be treated like insider risks. He listed observability principles including model diversity and incident disclosure. He argued the most trustworthy system will trust the model the least. Other tech leaders have also called for stronger safeguards. Some have urged pacing frontier development. President Trump opposes slowing US AI development. Trump launched a safety-focused task force for AI.