Microsoft president urges business to balance AI promise with safety and law

Brad Smith, president of Microsoft, told a United Nations Girls' Lounge audience that rapid advances in artificial intelligence raise both vast promise and significant peril, and that companies must take responsibility for safety.

"We all have an individual responsibility to produce products that are safe," Smith said, arguing that business leaders should ground decisions in international law and human‑rights norms and consider the broader social impacts of technology. He noted that markets can scale technology widely but sometimes fail to reach everyone: according to Smith, a recent report shows roughly 29% of people in the global North use AI versus 16% in the global South.

Smith pointed to a Security Council session on AI convening company leaders from firms such as Anthropic and OpenAI, and said governance requires more than the companies that build models: "It's the cloud operators... it's a whole ecosystem." He argued that firms should be prepared to slow product development when safety concerns arise, noting Microsoft reallocated engineers to address cybersecurity in the past.

The moderator, Shervin Bryce Pease of SABC, framed the discussion around whether corporations should change how they operate in light of AI's growth; Smith answered that firms, governments and third parties all have roles to play and that checks and balances—domestic regulation and multilateral cooperation—are necessary.