Brad Smith, Microsoft President, urged international cooperation on artificial intelligence safety regulations during a speech in Washington D. C. on October 12, 2026. He emphasized that AI systems, like everyday products such as milk and washing machines, require standardized safety checks to protect users worldwide. Smith argued that without unified rules, risks from advanced AI could grow unchecked across borders.
Smith drew a parallel between AI and common household items to illustrate his point. He noted that milk undergoes pasteurization and washing machines meet electrical safety standards before reaching consumers. Similarly, he said, AI models should pass rigorous testing for bias, reliability, and security before deployment. This analogy, he explained, makes the need for regulation more tangible to policymakers and the public. Smith stressed that innovation and safety are not opposing goals but complementary necessities in technological development.
The Microsoft president proposed a framework where nations agree on baseline safety protocols for high-risk AI applications. He suggested international bodies could oversee compliance, much like existing agreements for aviation or pharmaceuticals. Smith acknowledged challenges in aligning different legal systems but pointed to past successes in global cooperation on technology standards. He urged governments to act now, warning that delays could lead to fragmented rules that hinder both safety and innovation. Smith also highlighted Microsoft’s internal AI review processes as a model for responsible development.
Without coordinated action, Smith warned, companies might face conflicting requirements across countries, increasing costs and complexity. He cautioned that unsafe AI could erode public trust, slowing adoption of beneficial technologies in healthcare, education, and climate modeling. Conversely, he said, effective global rules could accelerate responsible AI growth by creating predictable conditions for businesses and researchers. Smith concluded that the coming years are critical for shaping AI’s trajectory, and leadership from major economies will determine whether the technology serves broad societal interests.
What specific AI applications did Brad Smith identify as needing regulation? Smith focused on high-risk systems used in areas like hiring, law enforcement, and critical infrastructure, where errors could cause significant harm to individuals or communities.
How does Microsoft currently approach AI safety internally? The company employs a responsible AI standard that includes impact assessments, testing for fairness, and ongoing monitoring of deployed systems across its products and cloud services.
Why does Smith believe global cooperation is achievable despite political differences? He cited historical examples such as international telecommunications standards and climate agreements, where nations found common ground on technical and safety issues despite broader disagreements.