Yoshua Bengio, founder of Mila, is co-chairing the first report from a new UN Scientific Panel on AI, a collaborative effort involving 40 experts from 37 countries working independently of governments or corporations. The Panel’s preliminary findings reveal a widening gap between rapidly advancing AI capabilities and the ability of policymakers to effectively govern them, highlighting a critical need for evidence-based understanding. UN Secretary-General António Guterres stated leaders “can no longer claim otherwise” regarding available scientific data on AI, framing the report as removing a key excuse for inaction. This assessment, mandated by UN General Assembly resolution, marks the beginning of an ongoing effort to provide independent, science-based evaluations of AI’s opportunities and risks and will be presented to global delegations this week in Geneva.
UN Panel Report: Assessing AI Capabilities and Safeguard Gaps
The rapidly accelerating development of artificial intelligence is outpacing humanity’s ability to understand and regulate it, according to a newly released report from a UN-convened panel of independent experts. Developed by 40 specialists from 37 countries working in their personal capacities, this was a deliberate effort to avoid national or corporate bias, and the assessment establishes a baseline understanding of AI’s capabilities and the widening gap between technological advancement and effective governance. This composition underscores the report’s intent to provide truly independent, international scrutiny of a field often dominated by commercial interests. Established last year through a UN General Assembly resolution, the Panel’s mandate centers on providing evidence-based insights to inform international discussions on AI.
The preliminary report, unveiled at a press conference in New York, highlights a critical finding: current safeguards are insufficient to manage the escalating power of AI systems. The report states that “Public policymakers cannot govern what they do not understand,” signaling a pressing need for increased scientific literacy among those responsible for shaping AI policy. The Panel deliberately refrained from offering specific policy recommendations, a strategic decision to preserve its scientific objectivity and prevent politicization of the findings. While acknowledging the limitations of current evidence, particularly regarding AI’s macroeconomic effects and environmental impact, the report aims to provide a common foundation of knowledge for governments worldwide. The Panel’s findings will be presented this week at the UN Global Dialogue on AI Governance in Geneva, offering a shared scientific starting point for international negotiations, and will continue to evolve through ongoing research and annual reports, with the next due in May 2027.
scientific data is now available and that leaders can no longer claim otherwise, adding that from now on, everyone must be part of the solution.
Mila’s involvement reflects a commitment to applying rigorous scientific methods to the global governance of AI, ensuring independent research informs its development. The panel comprised 40 experts from 37 countries, all contributing in their individual capacities, deliberately avoiding affiliations with governments, corporations, or institutions to maintain objectivity. This statement frames the report as removing a key justification for inaction on AI governance. Instead, the report serves as a foundational baseline of evidence for governments worldwide, even those not directly involved in AI development, to facilitate informed decision-making.
