Microsoft president says UK is best placed to drive AI regulation

We have reached a strange point in the tech cycle where the biggest players in the room are begging the government to slow them down. Microsoft President Brad Smith recently made a public push for the United Kingdom to take the lead on global AI regulation. His argument is simple: currently, there is more oversight on a carton of milk or a washing machine than there is on the most powerful technology in human history.

On the surface, it sounds responsible. It sounds like a big tech leader being a good citizen. But if you have been building in this space for more than a week, you know that when the largest incumbent in an industry starts asking for more rules, they are rarely doing it for the benefit of the small developer. They are doing it to build a moat.

The Milk Comparison and the Reality of Risk

Smith’s analogy about milk is effective because it appeals to common sense. We regulate food because it can kill you if it is spoiled. We regulate home appliances because they can start fires. The logic follows that we should regulate AI because of its potential for systemic harm. And he is right—AI has the potential to disrupt labor markets, spread misinformation, and compromise privacy at a scale we haven't seen before.

However, the comparison falls apart when you look at how innovation works. Milk is a finished product with a known chemical composition. Generative AI is a foundational technology. Regulating a foundation is much harder than regulating a byproduct. When you regulate the tool itself rather than the specific misuse of that tool, you end up stifling the guy in his garage long before you stop the bad actors.

Why the UK?

It is no coincidence that Microsoft is looking toward the UK. Post-Brexit, the UK has been trying to position itself as a "science superpower." They want to be the middle ground between the European Union’s heavy-handed, bureaucratic approach and the United States’ relatively hands-off, market-driven stance. By courting British regulators, Microsoft is trying to influence a framework that could become the international standard.

For builders, this is a double-edged sword. A single, clear set of rules is better than a fragmented global mess. If the UK can actually pull off a pro-innovation regulatory framework, it gives founders a predictable roadmap. But if they lean too far into the Microsoft-favored model, we might see compliance costs that only a trillion-dollar company can afford.

The Founder’s Perspective: Regulation as a Moat

Here is the honest truth from a founder’s perspective: regulation is often just another word for "barrier to entry." If the government mandates that every new model needs a multimillion-dollar safety audit before it can be deployed, who wins? Not the startup with a lean team and a great idea. Microsoft wins. Google wins. Meta wins. They have the legal departments and the capital to navigate the red tape.

We saw this with the early days of the internet. The companies that grew the fastest were the ones that operated in a gray area. Now that those companies have reached the top, they are more than happy to close the door behind them. By advocating for strict oversight now, Microsoft ensures that the next disruptive breakthrough has to happen within their ecosystem or not at all.

What Builders Should Watch For

If you are building an AI-native company right now, you need to pay attention to the specific language being used in these regulatory discussions. There are three key areas where these rules will hit you hardest:

• Liability: Who is responsible when a model hallucinates or provides dangerous advice? If the liability rests solely on the developer, small teams will be sued out of existence.
• Transparency Requirements: If you are forced to disclose your training sets and weights to the government, you lose your competitive advantage.
• Safety Audits: Third-party audits sound good, but if the auditors are expensive and the process takes six months, your speed-to-market is dead.

Smith is calling for "urgent" action, but urgency usually leads to poorly thought-out laws. We saw this with GDPR—a well-intentioned privacy law that mostly resulted in annoying cookie banners and a massive headache for small businesses, while the big players just hired more lawyers.

The Skeptic’s Takeaway

I don't doubt that Brad Smith believes AI needs guardrails. He is a smart guy, and he sees the risks. But we have to be skeptical of the timing. The call for regulation is happening just as open-source models are starting to catch up to the proprietary models owned by Microsoft and OpenAI. The more expensive it is to play the game, the fewer players there will be.

We need rules, but we need rules that focus on how AI is used in specific industries—like healthcare or finance—rather than broad, sweeping mandates on the code itself. Regulating the math behind the model is a dangerous path that leads to stagnation.

The goal should be to protect the public without protecting the incumbents. Right now, it looks like we might be headed for a world where only the giants are allowed to innovate.

For those of us in the trenches, the best move is to keep building fast and keep our code as modular as possible. Regulation is coming, whether we like it or not. The trick will be staying agile enough to pivot when the red tape finally arrives. Don't let the talk of "milk and washing machines" distract you from the fact that this is a battle for the future of who gets to build the next generation of tech.

Read the original at TechCrunch AI →