Microsoft Chairman and Chief Executive Officer Satya Nadella on 17 September 2026 called for artificial intelligence to remain firmly under human authority, warning that the technology was advancing faster than the scientific understanding needed to govern it. Speaking at the Ideas4India forum in Seattle, Nadella urged governments and technology companies worldwide to establish shared international norms, identifying a foundational understanding between the United States and China as an essential first step.

The Alignment Gap Nadella Named

Nadella was candid about the limits of the current moment. 'We do not yet have the science of alignment, and that's a fact,' he said. 'And yet we have powerful technology.' His framing drew a direct line between that gap and the urgency of governance: without alignment science, the only dependable safeguard is sustained human oversight.

He articulated his guiding principle plainly — 'any technology that doesn't serve humans or is not in control of humans is not worth doing.' To illustrate the dilemma facing the industry, he reached for a familiar analogy from software development: when engineers discover a critical flaw, should they pause the product or proceed? Nadella suggested the AI sector is navigating precisely that kind of inflection point, even while acknowledging that the competitive and societal forces behind AI development make an absolute halt unlikely.

The Case for International Governance

During an interactive session at the forum, Nadella returned repeatedly to the role of governments. He argued that any credible global governance framework had to begin with agreed norms between Washington and Beijing, but stressed that such a bilateral dialogue could not exclude the broader community of nations. His preferred model involved nation-states and AI-developing companies arriving at shared standards together.

Notably, he stopped short of proposing a specific negotiating architecture. His remarks were oriented toward first principles: that governments and industry must cooperate, and that the global AI race requires mutual understanding rather than pure strategic competition.

Data Sovereignty and Inclusive Participation

Nadella also pushed into the question of who controls the knowledge that AI systems generate. He argued that sovereignty mattered at the level of individual companies as much as countries. 'Countries need to have control of their knowledge,' he said, linking that protection directly to the conditions for broad participation in AI's benefits.

His argument was that genuine global engagement with AI — both contributing to and benefiting from it — would only occur when no country or company felt forced to surrender control of its data and intellectual assets to operate on someone else's technological frontier. 'It will only happen when all of the world can both benefit from and contribute to this knowledge,' he said.

He also cautioned against treating observed AI performance improvements as an inevitable, permanent law of nature. Drawing a comparison with Moore's law, he described current scaling laws as empirical findings that had held so far — not guaranteed projections.

India's Ambassador Weighs In

India's Ambassador to the United States, Vinay Kwatra, separately addressed the gathering, identifying decisions about data and its use in model training as matters for government, while affirming that the corporate sector carried significant responsibilities in India's AI development. Kwatra reaffirmed that India's AI mission should remain anchored in inclusivity, openness, sovereignty, democratised access, safety, and trust.

The Seattle forum followed the India AI Impact Summit 2026 and convened more than 40 Indian-origin chief executives and chief technology officers from fields spanning AI, quantum computing, semiconductors, cloud infrastructure, aerospace, and enterprise software. The event was convened by the Consulate General of India and also covered skills development and deeper India-US research and industrial cooperation.

With geopolitical competition over AI intensifying and alignment science still nascent, the pressure on governments and industry to move from principle to framework is only set to grow.