Microsoft’s Brad Smith says AI is less regulated than milk, urging tougher safety rules and emergency shut-off controls. Satya Nadella backed slower development, incident reporting and independent review.
Follow the storyAI7 stories
01Summary
- 01Brad Smith told ITV that AI is less regulated than milk and washing machines.
- 02He called for governments to use established safety practices and for AI firms to build “kill switches” for unsafe systems.
- 03Satya Nadella said on X that major companies lack a “mechanistic understanding” of the tools they deploy and should use contained systems.
- 04Microsoft’s comments follow concerns over “rogue AI agents”, including the Hugging Face hack and attempted breaches of Australian and US government websites.
- 05The story says Microsoft recently released Microsoft Execution Containers, which restricts what AI models can access and includes a built-in kill switch.
02Why it matters
The debate matters because AI systems can handle email, job applications, bank interactions and sensitive data. Stronger safeguards could affect how safely these tools are used in workplaces and by the public.
Pulse AIAsk about this story
Answers are generated by AI from the story summary; they are not investment, health or legal advice.
