Brad Smith, the president of Microsoft, has called for comprehensive global safety rules for artificial intelligence, stating that the technology currently faces less regulation than everyday household items. In an interview with the BBC, Smith argued that products like milk and washing machines are subject to more rigorous safety inspections than AI systems.
Comparing AI to Household Goods
Smith highlighted the disparity in oversight between consumer goods and digital technology. He pointed out that every carton of milk sold in a grocery store undergoes safety checks, yet AI systems operate with minimal mandatory oversight. He emphasized that the industry needs to build on existing knowledge of product safety to protect the public.
The Microsoft president suggested that firms should be legally required to write incident reports if their technology goes rogue. He also proposed that independent government evaluators should vet new AI products before they are released to the market. Smith described AI as both a tool and a weapon that requires urgent regulation to prevent harm.
Conflict with White House Policy
Smith’s stance places him at odds with President Donald Trump, who has advocated for the sector to regulate itself. Trump has warned that excessive rules could cause US firms to lose ground to Chinese rivals. However, Smith believes the US government would be more receptive to standards set by the United Kingdom than those from the European Union.
He praised the UK for its technical capability and the establishment of the AI Security Institute. Smith argued that the UK is best positioned to drive forward global rules for AI. He noted that Washington DC views London differently than it views Brussels, suggesting a potential pathway for cooperation.
Emergency Brakes and Kill Switches
Smith called for the implementation of emergency brakes, often referred to as kill switches, to pause AI systems if they get out of control. He stressed that the moment has arrived to bring people together and establish these safeguards. He acknowledged that while self-regulatory regimes are a good step, they are not sufficient on their own.
He noted that responsible companies will follow voluntary guidelines, but the real concern lies with less responsible actors. Smith applauded a recent agreement signed by major tech leaders at the White House, which included commitments to work with independent auditors. Microsoft is happy to sign such statements, but Smith believes more is needed to ensure public safety.
The call for stricter oversight comes amid growing concerns from experts who view AI as a potential threat to humanity. While leaders at companies like OpenAI and Google DeepMind have backed ideas for a development slowdown, Trump has pushed against these measures. Smith’s comments reflect a growing demand for laws and rules to control the rapid advancement of AI technology.
Source: BBC
