GENEVA — A United Nations independent scientific panel has urged governments to accelerate international cooperation on artificial intelligence governance, warning that the technology is advancing faster than scientific understanding and regulatory systems can keep pace.
The warning comes in the panel's first global scientific assessment of artificial intelligence, released ahead of the inaugural UN Global Dialogue on AI Governance scheduled for July 6-7 in Geneva. The report, prepared by the 40-member Independent International Scientific Panel on Artificial Intelligence, examines AI's opportunities, risks and impacts across economic, social, security and environmental sectors.
UN Secretary-General António Guterres said governments should act without delay to establish shared governance frameworks for AI.
"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," Guterres said at the report's launch. He added that the longer AI develops without internationally shared rules, the less influence governments and societies will have over its future direction.
The panel said AI has the potential to accelerate progress in fields including healthcare, education, agriculture and scientific research, but warned that increasingly capable systems also present growing risks. Those include misinformation, cyber threats, impacts on democratic institutions, widening inequality, environmental costs and concerns over the reliability and control of advanced AI systems.
According to the report, policymakers face an "evidence challenge" because AI capabilities are evolving more rapidly than governments can generate the scientific evidence needed to regulate them effectively. The panel said existing governance mechanisms remain fragmented and often rely on limited or company-provided information when assessing advanced AI systems.
The scientific body stressed that it does not set policy or regulations. Instead, it aims to provide an independent evidence base to help governments make informed decisions as AI technologies continue to develop. Panel members serve in their personal capacities and are independent of governments, companies and other institutions.
The preliminary assessment identifies seven areas for continued study, including AI's technological trajectory, economic effects, human rights implications, governance, security and societal applications. A more comprehensive report is expected in 2027 following additional consultations and research.
The findings will form the scientific basis for discussions among UN member states at the Global Dialogue on AI Governance in Geneva, where governments are expected to consider approaches to strengthening international cooperation on the rapidly evolving technology.
