Key information
Microsoft president Brad Smith recently argued that artificial intelligence currently faces fewer safety regulations than common household items like washing machines or milk. Speaking on the BBC's Big Boss podcast, Smith advocated for global safety standards, noting that existing oversight for food and utilities is more rigorous than current AI governance.
Smith proposed several specific safety measures, including a legal requirement to notify victims if systems are compromised by rogue AI agents. He also suggested that autonomous systems should feature a safety brake and undergo testing by outside evaluators prior to public release, characterizing AI as both a tool and a potential weapon.
These recommendations contrast with the current US administration's focus on maintaining technological dominance. While President Trump has issued executive orders to promote AI development, Smith is looking toward the United Kingdom for leadership, citing the nation's 2023 establishment of the AI Security Institute as a positive model for state-backed safety oversight.
Other industry leaders, including Anthropic's Dario Amodei, have also called for slower development and independent safety verification. Although major firms like Google and Nvidia have agreed to work with auditors under the White House Accord on Super Intelligence, Smith remains cautious, warning that self-regulation may not be sufficient to address the risks posed by less responsible companies.
