GENEVA—Governments, technology companies and international organisations on Monday advanced discussions on artificial intelligence governance frameworks ahead of the United Nations’ Global Dialogue on AI Governance, as officials warned that regulatory approaches are struggling to keep pace with rapidly evolving systems.
The two-day dialogue, convened by the UN General Assembly and coordinated with agencies including the International Telecommunication Union (ITU), opened in Geneva on July 6 with participation from all 193 UN Member States alongside private sector representatives, academia and civil society, according to UN programme documents and officials.
UN Secretary-General António Guterres, in opening remarks delivered at the session, said AI development was advancing “at runaway speed” and stressed that global coordination was necessary to ensure safety, fairness and accessibility. He added that the forum aimed to ensure “every country has a seat at the table,” reflecting concerns that existing governance debates have been dominated by a small number of technologically advanced states.
The dialogue builds on a UN General Assembly mandate adopted in 2025 establishing the Global Dialogue on AI Governance as a permanent multistakeholder platform for international cooperation on artificial intelligence, and follows months of consultations among member states and stakeholders.
The discussions come as an independent UN scientific panel on artificial intelligence released its first preliminary assessment ahead of the meeting, warning that AI systems are developing faster than regulatory and scientific understanding in many jurisdictions. The panel, composed of 40 international experts, highlighted risks including uneven global access to AI capabilities, governance gaps in oversight mechanisms, and the potential misuse of advanced systems.
Officials involved in the panel process said the findings were intended to inform international dialogue rather than prescribe binding rules, underscoring the non-regulatory nature of the scientific body.
The UN has also emphasised that disparities in infrastructure and computing capacity could widen inequality between countries, with some relying heavily on external AI systems while lacking domestic governance control, according to recent panel analysis circulated ahead of the forum.
Diplomatic sources said participating delegations were focusing on early-stage convergence around shared principles such as transparency, human oversight and accountability, though differences remain over enforcement mechanisms and the extent of international coordination.
The Geneva meeting is expected to conclude on July 7, after which organisers will compile inputs into a structured set of thematic discussions feeding into subsequent UN sessions on AI governance scheduled for 2027.
