He urged the international community to build a future “where policies are as smart as the technology they seek to guide.”
New panel of experts
The Secretary General noted that “innovation in AI is advancing at the speed of light, outpacing our collective ability to fully understand it, let alone govern it.”
He highlighted that “If we want AI to serve humanity, policies cannot be based on guesswork”, underscoring the need for “facts we can trust (and share) across countries and sectors.”
For this reason, the UN is developing mechanisms that put science at the center of international cooperation on AI, starting with a newly appointed body that brings together 40 leading experts in the field.
The Independent International Scientific Panel on Artificial Intelligence aims to help close “the AI knowledge gap” and assess the real impacts that these new technologies have on economies and societies so that countries can act with the same clarity regardless of their level of AI capability.
Accelerate progress, anticipate risks
“The Panel will provide a shared basis of analysis, helping Member States move from philosophical debates to technical coordination; and anchor options in evidence,” he said.
The UN chief insisted that science-based AI governance “is not a brake on progress” but rather “a solutions accelerator.”
It will help countries identify where AI “can do the most good, fastest,” he said, and provide “a way to make progress safer, fairer and more widely shared.”
In addition, the international community will be able to anticipate early the impacts of AI, such as risks for children or labor markets. That way “countries can prepare, protect and invest in people.”
Fragmentation Hazards
He noted that international cooperation is difficult today amid strained trust and growing technological rivalry.
“Without a common foundation, fragmentation wins, with different regions operating under incompatible policies and technical standards,” he said, which will only “increase costs, weaken security and widen divisions.”
The Secretary-General said countries can align their “technical baselines,” guided by the Independent Panel and another UN initiative, the Global Dialogue on AI Governance, to be held in Geneva in May.
Meaningful human supervision
Before concluding, he argued that while “science informs,” human control of AI must be “a technical reality, not a slogan.”
This requires “meaningful human oversight in every high-stakes decision – in justice, healthcare, credit” as well as “clear accountability – so that responsibility is never outsourced to an algorithm” said.
“People need to understand how decisions are made, question them and get answers.”