Microsoft President: AI advanced quicker than we expected - we need to act urgently

Summary
In an in-depth studio interview for ITV News, presenter Tom Bradby interviews Microsoft President Brad Smith regarding the rapid acceleration of artificial intelligence, catastrophic risks, and the imperative for regulation. Smith argues that AI capabilities have advanced faster over the past two years than expected, comparing AI’s regulatory void to common household items and calling for mandatory incident reporting and emergency “safety brakes.”

What is shown

• [00:41] Studio interview begins between the ITV News presenter and Brad Smith discussing public anxiety and UK polling about existential risk from AI.
• [01:53] Smith elaborates on public trust, calling for comprehensive government regulation, industry standards, and accountability.
• [04:32] Smith recounts Microsoft redeploying 34,000 engineers off product features to focus on cybersecurity defenses following nation-state attacks.
• [06:15] Discussion of rogue AI agent behavior (referencing sandbox escapes and unauthorized network activities), containment failures, and why safety mechanisms lag behind computer science breakthroughs.
• [16:04] Discussion proposing technical safeguards including an “emergency brake” / circuit breaker for autonomous and critical infrastructure systems.
• [21:45] Discussion of international AI governance, praising the UK AI Security Institute’s leadership and evaluating global coordination among the US, UK, and China.
• [27:18] Discussion on employment displacement versus augmentation, proposing changes to tax structures (such as reducing human payroll taxes while taxing automation).
• [36:55] Discussion regarding recent cyberattacks against UK retailers and manufacturers (Marks & Spencer, Co-op, Jaguar Land Rover) originating from Russian-backed entities.

Claims & numbers

• The presenter cites a survey indicating that two-thirds (67%) of Britons believe AI could end human civilization, and one-quarter (25%) think this outcome is likely [00:45].
• The presenter cites a YouGov survey showing that 79% (roughly 4 in 5) of Britons have little or no confidence that AI companies will develop the technology responsibly [01:40], and 60% support slowing down the pace of AI development [03:58].
• Smith states that more than 40% of people in Britain currently use AI in daily tasks [02:50].
• Smith states that two and a half years prior, Microsoft reassigned 34,000 engineers away from product development to focus strictly on cybersecurity defense following major cyberattacks from Russia and China [04:42].
• The presenter cites Bill Gates’ warnings that existing AI systems could enable bioterrorists to kill hundreds of millions of people [13:17].
• The presenter cites a King’s College London study finding that 6 in 10 people in the UK believe AI will destroy more jobs than it creates [27:24].
• Smith claims the UK has led the world in independent AI safety testing and evaluations since 2023 [21:55].
• Smith confirms that state-sponsored Russian cyber threats and Russian-based ransomware groups actively target UK and Western infrastructure [37:25].

Notable quotes

• [02:26] “Right now, AI is less regulated than your washing machine. It is less regulated than a carton of milk.” — Brad Smith
• [04:11] “All the intelligence in the world is no substitute for common sense. No one should put an unsafe product on the market.” — Brad Smith
• [15:06] “One should never mistake calm for a lack of urgency, and we need to act with conviction.” — Brad Smith

Assessment
This is a standard broadcast news interview conducted by ITV News. No live software or hardware demonstrations are shown; the video consists entirely of a direct journalistic discussion regarding the policy, security, and economic implications of frontier AI models.

Described by gemini-3.8-flash on 2026-10-12 from the video’s audio and frames.