Microsoft President Brad Smith recently called for stronger regulation of artificial intelligence (AI), warning that current oversight of AI is even weaker than that applied to milk and washing machines, according to foreign media outlets including the BBC and The Guardian on the 12th. In addition, Microsoft founder Bill Gates also called for stronger AI regulation in an interview with US media on the 11th, saying that this "most dangerous technology humans have ever invented" cannot be left entirely in the hands of the companies developing it. Gates said that allowing tech giants to self-regulate could expose the world to unprecedented cyber threats.

"Right now, your washing machine is subject to more safety rules than AI, and every carton of milk you see in the grocery store must comply with safety rules and be inspected," Smith said in a recent interview on a BBC program, noting that AI is both a "tool" and a "weapon" and urgently needs regulation. Smith argued that global safety rules need to be established, that companies need to be equipped with an "emergency brake" to pause or slow down AI operations when something goes out of control, and that relevant parties need to bear legal obligations toward the "victims" of runaway AI agents.

Russia Today (RT) reported on the 12th that Bill Gates also said in an interview with CNN, "For the most dangerous technology humans have ever invented, if you think this is the final word, that society's only check is just the companies' own regulation, then that is clearly not enough." Gates said government agencies must play a more active role, and he also compared AI to the aviation and pharmaceutical industries, where regulators enforce safety standards.

Gates also identified cyberattacks as the most immediate threat, warning that criminals could exploit AI models with hacker and bioterrorism safeguards removed. "When bad actors, systems without safety safeguards, and the latest models come together... then in the coming year, we will face many unexpected negative consequences," Gates said.

The Guardian mentioned in its report that cases of AI "going out of control" are increasing. This summer, OpenAI and Anthropic separately disclosed that their AI models broke through testing environments into the public internet during cyber offense and defense capability tests and accessed other institutions' systems without authorization, raising concerns about the risk of AI "going out of control."

Discussion in the US AI industry has recently intensified around the pace of development, security risks, and government regulation. Previously, several US tech entrepreneurs called on AI companies to slow the development of frontier technologies. Dario Amodei, CEO of US AI company Anthropic, said in an article on September 12 that, considering safety risks, AI companies need to slow the pace of developing the most advanced models. OpenAI CEO Sam Altman and US entrepreneur Elon Musk subsequently said they support controlling the pace of AI model development.