The rapid concentration of artificial intelligence (AI) power is outpacing the world’s ability to understand and manage its risks.
So says United Nations (UN) secretary-general António Guterres, who is calling for stronger international safeguards and oversight of the technology.
Opening the General Debate of the 81st session of the UN General Assembly yesterday, in New York, Guterres said the balance of power had historically been shaped by territory, industry and finance, but is increasingly being determined by data, computing capacity and algorithms.
“This new power is highly-concentrated,” he said. “It is advancing at extraordinary speed and racing ahead of our ability to understand its full consequences.”
Guterres said he had warned almost a decade ago that AI could transform lives while also disrupting labour markets, global security and societies.
He argued that the technology itself is not the central threat, but rather the absence of accountability and oversight around increasingly powerful AI systems.
“The danger is not technology,” Guterres said. “The danger is technology without accountability: Capability without oversight. Decision-making without transparency.”
The UN secretary-general said the risks are becoming more difficult to ignore as AI capabilities expand, including among the companies developing the technology.
Taking greater responsibility
He called on governments, particularly those with the greatest AI capabilities, to take more responsibility for managing emerging risks and creating mechanisms for dialogue, transparency and cooperation.
Guterres compared the challenge posed by AI to the management of nuclear weapons during the Cold War, when geopolitical rivals established mechanisms intended to reduce misunderstandings, manage escalation and prevent catastrophe.
“Today, artificial intelligence demands the same clarity of purpose and sense of urgency,” he said.
He urged countries at the forefront of AI development to share information about emerging safety risks, cooperate on testing and evaluation, and work towards common safeguards.
Guterres also argued that AI governance could not be left to individual countries because the technology’s impact crossed national borders. “Because this affects us all, global coordination is indispensable.”
He pointed to existing UN initiatives, including the Global Digital Compact, which established universal principles for AI governance, as well as the Global Dialogue on AI Governance and the Independent International Scientific Panel on AI.
The scientific panel released a thematic brief on AI safety and risks associated with the loss of human control ahead of the General Assembly debate.
Guterres said the next phase should focus on what he described as the responsible pacing of AI development and additional safeguards to address emerging risks.
He also called for work towards a multilateral AI Risk Management Framework backed by independent oversight.
The UN chief stressed that AI governance should also address the global digital divide and ensure developing countries were not excluded from the benefits of the technology.
“Artificial intelligence must be shaped by humanity, with humanity and for all humanity,” he said.
Guterres singled out two areas where he said international consensus should be clear: children should not be exposed to unregulated AI systems, and machines should not be given control over life-and-death decisions.
“Children must never become the test subjects for unregulated systems. And let us resolve that life-and-death decisions must never be surrendered to machines.”
He also rejected the prospect of autonomous weapons becoming part of future warfare.
“Killer robots must have no place in our future,” Guterres said.
AI turns ‘superintelligence’
US president Donald Trump used his address to the UN General Assembly to frame AI as a key component of future economic and geopolitical competition.
Trump argued that the technology should be described as “superintelligence” rather than artificial intelligence.
“The use of the word artificial makes intelligence fake,” he said. “It is not fake. It’s actually amazing.”
Trump said technological leadership would be central to competition between countries in the years ahead.
“Whoever wins superintelligence wins,” he declared, while saying the US would encourage development of the technology, while monitoring potential risks.
The contrasting emphasis in the two speeches highlights a growing international debate over how AI should be governed as capabilities advance.
Guterres focused on international cooperation, risk management, transparency and safeguards, while Trump stressed technological leadership and continued development of advanced AI.
The debate comes as governments, technology companies and researchers grapple with the implications of increasingly capable AI systems, including their potential effects on employment, scientific research, security and decision-making.
For the UN secretary-general, the central challenge is ensuring the speed of technological development does not outstrip society’s ability to govern it.
The issue, he argued, is no longer simply whether AI will transform economies and societies, but whether governments can establish common rules and safeguards quickly enough to manage the consequences.
Share
