The United Nations has warned that artificial intelligence is advancing faster than science and regulation, leaving governments without enough safeguards to control the technology safely.
A new preliminary report by the UN’s Independent International Scientific Panel on Artificial Intelligence says AI capabilities are outpacing both scientific understanding and governments’ ability to adapt. The panel is made up of 40 scientists and experts from all five UN regions.
“AI capabilities are outpacing both scientific understanding and governments’ ability to adapt,” said Yoshua Bengio, co-chair of the panel. “With growing evidence of deceptive AI behaviour, science currently cannot guarantee that as capabilities continue to increase, AI will not cause catastrophic harm, either on its own or due to malicious users.”
UN Secretary-General António Guterres launched the assessment on July 2, 2026, and said the world can no longer claim ignorance about AI’s risks. “The science is here. We can no longer say we did not know. What we do with it is now up to all of us,” he said.
He added that without shared global standards, many governments and citizens may lose influence over how AI shapes society.
The report describes AI as having “enormous” potential benefits in health, education, agriculture, climate and security.
It also highlights serious risks, including harms to mental health, misuse as a destructive tool, and challenges in controlling highly autonomous systems.
“Sycophantic AI behaviour, where AI responses reinforce users’ existing beliefs regardless of accuracy, has been linked to several severe mental health incidents, including documented deaths,” the report stated.
Panel co-chair Maria Ressa said policymakers need evidence to make good decisions, but that evidence often comes too late because of AI’s speed.
The report notes that AI task complexity is doubling every four to seven months, and that over one billion people now use conversational AI weekly.
Development remains concentrated, with the United States accounting for 75 percent of computing power among the world’s top 500 AI supercomputers and China 15 percent.
The panel expects a shift toward agentic AI systems that can carry out real-world tasks with little or no human oversight, with implications for labour markets, cyber security and scientific research. It warns that the rapid, unchecked deployment of the technology at scale also presents considerable risks.
Guterres urged urgent global cooperation before the inaugural Global Dialogue on AI Governance in Geneva. “If AI continues to advance without shared global standards, many governments and citizens across the world may lose influence over how the technology may shape the society,” he said.
