UN launches independent global scientific panel on AI to close the knowledge gap
The UN has launched the Independent International Scientific Panel on Artificial Intelligence and nominated 40 experts to serve on it. Framed as the first global, fully independent scientific body focused on AI, the panel is tasked with closing the knowledge gap and assessing AI's real impacts across economies and societies.
"AI is moving at the speed of light. No country can see the full picture alone," said UN Secretary-General Antonio Guterres. He added the panel will "help the world separate fact from fakes, and science from slop … at a moment when reliable, unbiased understanding of AI has never been more critical."
What the panel will do
- Provide authoritative, science-based assessments of AI's impacts across sectors and regions.
- Act as a global reference point for policymakers and the public-separating evidence from misinformation.
- Support shared guardrails and international cooperation while encouraging innovation for the common good.
- Operate on an accelerated timeline, with its first report due ahead of a Global Dialogue on AI Governance in July.
Who's on it
The proposed 40 experts were selected from more than 2,600 applicants worldwide. Their expertise spans machine learning, data governance, public health, cybersecurity, child development, and human rights.
Members are expected to serve in a personal capacity-independent of governments, companies, or other institutions-to keep the panel's work objective and credible.
Why this matters for scientists and research leaders
Policy efforts on AI remain fragmented while deployment races ahead. A common scientific baseline can reduce noise, align methodologies, and focus resources on what actually works and what clearly doesn't.
- Methods: Expect pressure for transparent evaluation protocols, domain-specific benchmarks, and reproducible evidence.
- Risk: Increased attention on systemic risks (incidents, misuse, security, bias) and measurable mitigation strategies.
- Society: Greater weight on health, education, labor markets, child well-being, and rights-based impacts.
- Data: Calls for higher data integrity, documentation, and governance that support cross-country comparability.
- Translation: Findings that bridge lab results and real-world outcomes, informing standards and procurement.
Timeline and what to watch
The panel's first report is expected before the Global Dialogue on AI Governance in July. That timeline means near-term influence on how governments, standards bodies, and funders set priorities.
- Structure of the first report: Scope, definitions, and what counts as reliable evidence.
- Evaluation: Recommendations on benchmarks, audits, and incident reporting across domains.
- Safeguards: Practical guardrails for high-risk uses, including measurement plans.
- Coordination: How the panel proposes to work with national agencies, standards groups, and existing initiatives.
How research teams can prepare
- Document rigor: Publish clear measurement plans, datasets, model cards, and reproducibility artifacts.
- Track outcomes: Pair technical metrics with socioeconomic and safety outcomes from deployments.
- Contribute evidence: Aggregate incident data, error analyses, and ablation studies that speak to real-world risk.
- Propose testbeds: Offer domain-specific benchmarks or evaluation environments that reflect operational constraints.
- Engage early: If consultations open, submit concise briefs with methods, datasets, and results that can scale internationally.
The panel was mandated under the UN's Pact for the Future and established by a UN General Assembly resolution. For context on the UN's AI agenda and the Pact, see the UN's pages on Artificial Intelligence and the Pact for the Future.
If you're building the skills base within your lab or institute, you can explore curated AI course paths by role here: Complete AI Training - Courses by Job.
Your membership also unlocks:
