Mustafa Suleyman: testing 10x-larger models requires removing guardrails

Microsoft AI chief Mustafa Suleyman told Bloomberg about recent AI safety incidents, the risks of removing guardrails while testing future models 10x larger than today's, and the need for a cross-industry safety body. He also said the government should help "drive" the process of evaluating models.

• Suleyman: testing 10x-larger models requires removing guardrails
• Microsoft AI chief backs a cross-industry AI safety body
• He says government should help "drive" model evaluation
• Interview touches on recent AI safety incidents