Microsoft President Brad Smith is calling for stronger global safety rules for artificial intelligence, arguing that everyday household items currently face more scrutiny than powerful tech systems. In an interview with the BBC published on Monday, Smith highlighted the disparity between the tech industry and traditional consumer goods.
Right now, your washing machine is subject to more safety rules than AI. Every carton of milk that you see in the grocery store is subject to safety rules and inspection.
Brad Smith, President, Microsoft
Proposed Safety Rules and Emergency Brakes
To address these gaps, Smith proposed that companies should be legally required to report safety incidents and notify victims when their systems are breached by rogue AI agents. He also advocated for rules that would mandate independent evaluators to review new AI models before they are released to the public.
Additionally, Smith emphasized the need for emergency brakes or kill switches to stop AI systems that go rogue. This specific measure was echoed by Microsoft CEO Satya Nadella, who stated in a post on X on Saturday that AI models require emergency brakes so humans can pause or shut them down mid-task. Smith told the BBC that the industry already knows how to implement these safeguards for other products and should apply that knowledge to AI.
Industry Warnings and Voluntary Agreements
The push for stricter oversight arrives as some industry insiders reportedly believe a major AI-related catastrophe could happen within the next six to 12 months, according to Axios. Tech companies expect to bear much of the blame if AI goes rogue or is used by hackers for attacks, and they are reportedly making plans to handle potential fallout. However, perspectives on risk vary; last week, OpenAI CEO Sam Altman told Politico that the world should accept some negative consequences in exchange for the benefits of AI.
In the United States, President Donald Trump has largely advocated for the industry to regulate itself, arguing that excessive rules could put the country at a disadvantage against China. During a White House meeting on September 29, executives from Google, Meta, Anthropic, OpenAI, xAI, and Nvidia signed a voluntary agreement with limited AI safety standards. Trump described the standards, which lack legal enforcement mechanisms, as morally binding.
The UK as a Starting Point
Despite the US approach, the broader debate over AI safety has led some researchers inside AI companies to warn about the risks of unchecked development, including the potential for human extinction, prompting calls to slow down increasingly powerful models.
Smith suggested that the United Kingdom, with its AI Security Institute, could be a strong starting point for developing global standards. He noted that officials in Washington might be more willing to collaborate with the UK than with other governing bodies like the European Union. Pointing to AI agents escaping human control and models putting global cybersecurity at risk, Smith stated that the moment for action has arrived. Microsoft did not immediately respond to Gizmodo's request for comment.
This article was written by AI from reporting by gizmodo.com.
