
The Interview Brad Smith, Microsoft President: Washing machines are better-regulated than AI
Oct 30, 2026
Brad Smith, Microsoft’s president and vice-chairman, has spent decades navigating technology, law and public policy. He discusses why AI needs safety standards before a crisis, including incident reporting, independent evaluations and government certification. He also explores the limits of self-regulation, the community impact of data centers, and how tech companies can work with governments while protecting innovation.
AI Snips
Chapters
Transcript
Episode notes
Treat AI As Both Tool And Weapon
- AI must be treated as both a tool and a weapon, with safeguards addressing harms before they affect an entire generation.
- Smith says delayed action cannot be recovered because Microsoft has no time machine; regulators should begin today.
Build Three Safety Layers Into AI
- Require AI systems controlling autonomous devices or critical infrastructure to include emergency brakes, incident reporting, and independent safety evaluation.
- Smith compares these safeguards with household circuit breakers, privacy-breach notifications, and aircraft certification.
Pair Industry Standards With Government Rules
- Effective AI regulation may combine industry standards with legally binding government rules rather than relying exclusively on either side.
- Smith identifies the UK as unusually well positioned because its AI Safety Institute combines technical capability with international credibility.

