UN Creates Independent Scientific Panel on AI Governance
The United Nations has established a 40-member Independent International Scientific Panel on AI to assess artificial intelligence development through an evidence-based lens. The UN General Assembly formalized the panel through resolution A/RES/79/325 in August 2025, setting its mandate and requiring members to serve in personal capacities independent of governments, companies, and institutions.
The panel originated from the Global Digital Compact, adopted at the UN's 2024 Summit of the Future. That agreement committed member states to creating both an independent scientific body and a Global Dialogue on AI Governance-two mechanisms designed to ground international AI policy in evidence rather than assumption.
Why This Matters for Biomedical Research
For biomedical researchers, the panel's creation addresses a specific problem: AI innovation is advancing faster than clinicians, regulators, and patients can evaluate it. Independent scientific assessment has become essential to realizing AI's potential in healthcare and other domains.
The panel includes researchers and experts across machine learning, ethics, governance, and domain-specific fields. Co-chairs Maria Ressa and Yoshua Bengio lead the group alongside 38 other members drawn from academic institutions, research organizations, and independent positions worldwide.
Structural Independence
The UN Office for Digital and Emerging Technologies houses the panel's secretariat. All members disclosed financial, professional, and personal interests to prevent conflicts of interest-a requirement designed to preserve the panel's impartiality when advising on AI policy.
The panel's dual-track approach-pairing scientific assessment with the Global Dialogue on AI Governance-creates a feedback loop between evidence and policy deliberation. Scientists provide technical evaluation while policymakers and stakeholders engage in broader governance discussions.
What's Next
The panel now faces the task of evaluating AI systems across sectors, from healthcare to criminal justice. For research professionals, this creates both an accountability structure and a resource: independent assessment of which AI tools warrant adoption and which require further scrutiny.
Learn more about AI for Science & Research or explore AI Research Courses to understand how scientific evaluation shapes AI deployment.
Your membership also unlocks:
