Microsoft president calls for AI safety rules and help for workers

Microsoft president Brad Smith says governments should prepare for AI-related job losses and require stronger safety measures for advanced systems. His proposals range from support for workers to independent testing and an emergency switch-off mechanism, making this a substantial intervention in the debate over who should set the rules.

Microsoft AI Watch analysis

What happened

In an interview with ITV News, reported by Yahoo News UK on 8 October, Smith said AI will displace some jobs and governments will need to help people manage the transition. He raised retraining, taxing companies that replace workers with AI or robots, and potentially protecting roles such as teaching, medicine, judging and care work from being performed by AI.

Smith also called for independent safety evaluations before models launch, disclosure when AI agents go rogue or users are hacked, and an “emergency brake” that could shut systems down. He said governments should be able to order systems offline and argued that developers could face legal liability if negligence led to catastrophic harm.

Why it matters

These are not just general warnings about powerful technology. Smith is proposing specific obligations for developers, infrastructure providers and governments, while also acknowledging that AI deployment could disrupt work. The ideas touch both sides of the public bargain: who bears the costs of change, and who can intervene when systems go wrong.

Some proposals, including protecting particular professions, would raise difficult questions about what “performed by AI” means in practice. But the interview gives readers a concrete set of policy choices to debate rather than another foghorn call for responsible innovation.

Our read

Smith’s list is notable because it combines worker support with testing, reporting and the ability to switch systems off. The harder question is how to turn those principles into rules that are enforceable without making safety a slogan or a brake applied only after trouble starts. His proposals deserve scrutiny on their details, not just applause for sounding sensible.

What to watch

• Whether Microsoft turns these proposals into specific policy recommendations.
• How governments approach independent evaluations and incident disclosure.
• Whether worker protections focus on retraining and income support, or restrict which jobs AI can do.

Discussion spark: Should governments be able to order AI systems offline and protect selected professions from automation, or should policy focus on worker support and safety testing instead?

not affiliated with or endorsed by Microsoft