Translated with AI
On February 20, 2026, United Nations Secretary-General António Guterres emphasized the key role of science in the global governance of artificial intelligence, speaking in New Delhi at an event held on the sidelines of the AI Impact Summit, according to the UN press service.
"By building on science, we can transform artificial intelligence from a source of uncertainty into a reliable driver for achieving the Sustainable Development Goals," he said.
The Secretary-General called on the international community to build a future "in which policies are as smart as the technologies they are designed to guide."
Guterres noted that AI innovations are advancing at a rapid pace, outpacing the collective capacity of countries to fully understand and effectively manage them. He argued that for AI to serve humanity, policies must not be based on guesswork. What's needed are "facts that can be trusted and shared across countries and sectors."
This is why the UN is developing mechanisms to place science at the center of international cooperation in AI. Specifically, the Independent International Scientific Panel on Artificial Intelligence, comprised of 40 leading experts, was recently established.
The group aims to bridge the AI knowledge gap and assess the real-world economic and societal impacts of new technologies, so that governments can make decisions with an equally clear understanding of the situation, regardless of their own technological advancement.
"The Group will provide a common analytical framework, helping Member States move from philosophical discussions to technical coordination and evidence-based decision-making," the UN chief emphasized.
He stated that science-based governance of AI will not hinder progress, but rather accelerate decision-making. This approach will allow countries to more quickly identify where technologies can bring the greatest benefit, as well as ensure that development is safer, fairer, and more accessible for all.
Furthermore, the international community will be able to identify in advance the potential impacts of AI use – for example, risks to children or the labor market – allowing countries to “prepare, protect people, and invest in their future.”
Guterres noted that international cooperation today is complicated by declining trust and increasing technological rivalry.
"Without a common framework, fragmentation takes over, with different regions operating under incompatible policies and technical standards," the UN chief warned. He added that this increases costs, reduces security, and deepens inequality.
The Secretary-General indicated that countries could agree on "technical benchmarks" based on the work of the International Panel and another UN initiative, the Global Dialogue on AI Governance, which will be held in Geneva in May.
Concluding his speech, he emphasized that human control over AI must be "a technical reality, not a slogan." This means the need for real human oversight in high-stakes decision-making—in justice, healthcare, and lending—as well as a clear accountability system to ensure that responsibility is never shifted to an algorithm.
“People need to understand how decisions are made, be able to challenge them and get answers,” the Secretary-General concluded.
