UN Panel Warns AI Governance Is Falling Behind Rapid Technological Advances

A United Nations-backed scientific panel has warned that global safeguards governing artificial intelligence are failing to keep pace with the technology’s rapid development, raising concerns over growing risks and unequal access to AI across the world.

The warning came on Wednesday as the Independent International Scientific Panel on AI released its first global assessment examining both the opportunities and challenges presented by artificial intelligence. The report will serve as a key reference during the inaugural Global Dialogue on AI Governance, scheduled to take place in Geneva on July 6 and 7.

The 40-member panel was established by the UN General Assembly in August last year to provide independent scientific analysis of AI developments and their global implications.

Panel co-chair Yoshua Bengio said AI systems are advancing at a faster pace than governments and scientific research can fully understand or regulate. He pointed to increasing evidence that some advanced AI models have displayed deceptive behaviour during testing and warned that researchers cannot yet guarantee such systems will remain free from catastrophic risks, whether caused by technical failures or malicious misuse.

Co-chair Maria Ressa said the report identified three major trends shaping the future of artificial intelligence: rapidly expanding capabilities, growing concentration of power and declining human control. She noted that performance in “Humanity’s Last Exam,” a benchmark consisting of 2,500 expert-level questions across multiple disciplines, improved from 8 percent to 45 percent in just 16 months, highlighting the speed of AI progress.

The report also found that AI development remains heavily concentrated in a small number of countries and companies. According to the panel, the United States accounts for 75 percent of the computing power used in the world’s largest AI clusters. Private companies dominate the sector, producing 91 percent of notable AI models released in 2025. US institutions developed 59 major AI models during the year, compared with 35 from China and only 13 from all other countries combined.

UN Secretary-General Antonio Guterres said the assessment had been distributed to governments worldwide, urging policymakers to act before AI development outpaces public oversight. He warned that without internationally shared rules, governments and citizens would lose influence over how the technology evolves.

The panel also expressed concern that much of the Global South remains excluded from both AI development and decision-making, despite facing many of the technology’s future economic and social impacts. It said this imbalance risks widening existing global inequalities by leaving many countries without the resources or influence needed to shape AI governance.

While the report evaluates scientific evidence across areas including security, economics, environmental impacts, democracy, human rights and governance, the panel stressed that its role is to provide evidence rather than recommend specific policies. Decisions on international regulations will be left to UN member states during discussions at the Geneva dialogue and future negotiations.

The panel’s next comprehensive report is expected to be presented ahead of the second Global Dialogue on AI Governance, scheduled for New York in May 2027.