Microsoft President Brad Smith has called for urgent regulation of AI, arguing that consumer products like milk and washing machines are subject to greater scrutiny than companies that develop artificial intelligence.
In a recent BBC podcast, Mr Smith called AI a “weapon and a tool” that put global cybersecurity at risk, and said “the time has come” to set global safety rules.
Mr Smith also said the UK, rather than the US or the European Union, was best placed to lead on global safety rules. Founded in 2023, the UK’s AI Security Institute (AISI) quickly became a leader in AI safety testing, with key models available from frontier labs in the US prior to deployment.
That role has been under threat in recent weeks, with US President Donald Trump claiming that “globalist” organizations are interfering with American businesses. Last month, he reportedly asked Anthropic and OpenAI not to provide new models to UK testers, but AISI confirmed on Monday that the latest models were available to access and test.
But despite the president’s objections, Mr Smith insisted the UK could regulate AI in a way that the US cannot.
“In Washington DC, people don’t see Britain, and they don’t see London the way they see Brussels,” he told the BBC. “That’s a strength.”
His comments come amid calls from technology leaders, including Anthropic’s Dario Amodei and OpenAI’s Sam Altman, to slow the pace of AI development following several incidents in which AI agents have gone berserk.
Last month, President Trump invited technology leaders from companies such as OpenAI, Anthropic, Nvidia, AMD, SpaceX, Meta, and Google to the White House to discuss AI safety. They signed a “morally binding” document agreeing to work with “independent auditors” to evaluate the safety and functionality of the AI models.
Smith said the document is a “good step” but falls short of the kind of regulation needed. The problem with self-regulation, he says, is that only “good companies” comply.
“It’s the companies that are less responsible that we need to worry about,” he says.
Mr Smith outlined some AI safety regulations he would like to see introduced, including a legal obligation to warn organizations if their systems are compromised by rogue AI agents. He also seemed in favor of AI labs getting approval from external evaluators before AI products are released, and spoke positively about kill switches that would pause or slow down the AI if humans lose control.
“Let’s calm down, maybe speak gently but firmly, and start bringing people together,” Smith said. “I hope that’s the role we can play as a business. I hope that’s the role Britain can play as a country.”
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
