Microsoft President Brad Smith: AI Is Less Regulated Than Milk, and the UK Should Write the Rules

The Three-Question Framework

What Happened?

Brad Smith, president of Microsoft, told the BBC that AI is subject to fewer safety rules than a washing machine or a carton of milk. He asked for global rules that include "emergency brakes" to pause or slow AI, a legal duty to write incident reports and to tell a victim when a rogue AI agent breaches their system, and independent government evaluators who vet new AI before release.

Why It Matters

A major AI platform owner is publicly asking for binding rules. He also names the UK, through its AI Security Institute, as the best place to lead. That cuts against President Trump's position that the sector should regulate itself so US firms do not lose to Chinese rivals. Smith says self-regulation has a gap: "the good companies will do what you want" and the risk is the less responsible ones.

What Could Happen Next

Watch whether the UK moves to turn the AI Security Institute into a standards body, and whether Washington accepts UK-led rules. Smith argues the US would accept them more readily than rules from Brussels. A legal duty to notify victims is the concrete proposal most likely to be tested against real incidents, such as the Philadelphia case reported the same week.

Executive Key Takeaways

Smith: "Right now, your washing machine is subject to more safety rules than AI."

Three proposals: emergency brakes, mandatory incident reports with victim notice, and independent pre-release evaluators.

He backs the White House voluntary pledge but says it is "not sufficient any longer".

He says the UK is best placed to lead because of its AI Security Institute.

Trump has said the sector should regulate itself; Smith says Washington views London differently from Brussels.

Smith spoke in the BBC's Big Boss Interview, published on 11 October 2026. He described AI as both a "tool and a weapon" and said it needs regulating urgently: "Let's get started today."

His three concrete asks are specific. First, "emergency brakes", often called kill switches, so that firms can pause or slow an AI system that gets out of control. Second, a legal requirement to write incident reports and to tell a victim if a rogue agent breaches their system. Third, a rule on when outside evaluators must sign off on an AI product before release. He suggested that independent government evaluators should vet new AI technology.

The BBC puts these remarks next to the White House meeting at the end of September, where leaders of OpenAI, Anthropic, Nvidia, SpaceX, Meta and Google signed what it describes as a "morally binding" document. Under it the companies agreed to work with independent auditors to check that their systems work as intended and do not "hack or access technical systems in unintended ways". Smith said Microsoft is "happy to sign it" and called it "a good step, a necessary step", but not sufficient.

His reasoning for why a voluntary scheme falls short is short: "the good companies will do what you want", and "it's the companies that are not as responsible that you need to worry about."

Why the UK? Smith named the UK's AI Security Institute, launched under former prime minister Rishi Sunak as the AI Safety Institute, and said most people in the tech sector would name the UK as the leader in that technical capability. He also argued that Washington looks at London differently from Brussels, which in his view makes UK-written standards easier for the US to accept.

The BBC adds that the bosses of US-based AI developers Anthropic, OpenAI and Google DeepMind have backed the idea of slowing AI development, and that President Trump has pushed against that and has repeatedly said that "whoever wins AI, wins".

This report covers Smith's statements only. No draft law, UK government response or US response was part of the source.

The Three-Question Framework

What Happened?

Why It Matters

What Could Happen Next

Executive Key Takeaways

Primary Sources & Citations