Two computer scientists named to U.N. panel on artificial intelligence
Adji Bousso Dieng and Aleksandra Korolova, both on Princeton's computer science faculty, have been appointed to the United Nations' Independent International Scientific Panel on Artificial Intelligence. The panel will examine the risks, opportunities, and impacts of AI and translate technical evidence into guidance for policymakers.
Created in August 2025, the panel is described by the U.N. as the "first global scientific body dedicated entirely to artificial intelligence." It includes 40 international experts who will meet periodically over the next three years to produce assessments of the technology. According to the U.N., "The Scientific Panel on AI will serve as a crucial bridge between cutting-edge AI research and policymaking" and help the international community make informed decisions about governing this technology.
The panel will present an annual report at the United Nations Global Dialogue on AI Governance. Dieng and Korolova were selected from more than 2,600 applicants, nominated by Secretary-General António Guterres, and confirmed by the U.N. General Assembly on February 12.
Why this matters for researchers
Expect stronger demand for evidence that is measurable, reproducible, and relevant to policy. The panel's work will likely influence benchmarks for evaluation, expectations for responsible deployment, and how institutions assess risk.
- Measurement standards: Methods that quantify dataset diversity and model outputs (e.g., diversity scoring approaches such as the "Vendi Score") are poised to gain traction in evaluations and reporting.
- Governance and auditability: Algorithm and AI audits, fairness testing, and privacy guarantees will continue moving from research papers to standard practice.
- Public health and biosurveillance: Early-warning analytics for emerging variants highlight a growing link between machine learning and real-time scientific surveillance.
- Research cadence: Annual U.N. reports create clear checkpoints; align studies, benchmarks, and open data contributions so they're easy for policymakers to cite and act on.
About Adji Bousso Dieng
Dieng is an assistant professor of computer science whose research connects AI with the natural sciences. Her lab, Vertaix, introduced the "Vendi Score," a metric for assessing diversity in datasets and model outputs-useful for building dependable ML systems and accelerating scientific discovery. The team has applied this line of work to detect emerging variants of viral diseases like COVID-19 before they are formally identified.
Dieng joined Princeton in 2021 after serving as a research scientist at Google DeepMind (2020-2025). Her honors include the Prix Galien Africa Special Prize, a 2022 Schmidt Sciences AI2050 Early Career Fellowship, and the 2022 Annie T. Randall Innovator Award from the American Statistical Association. She holds a Diplôme d'Ingénieur from Télécom Paris, a master's degree from Cornell University, and a Ph.D. from Columbia University. At Princeton, she is affiliated with Chemical and Biological Engineering, the Princeton Materials Institute, Princeton Precision Health, the Princeton Quantum Initiative, the Andlinger Center for Energy and the Environment, and the High Meadows Environmental Institute.
About Aleksandra Korolova
Korolova is an assistant professor of computer science and public affairs whose work focuses on the societal impacts of AI. She develops and deploys algorithms that enable data-driven innovation while preserving privacy, fairness, and reliability, and she designs and conducts algorithm and AI audits.
Korolova joined Princeton in 2022 with a joint appointment in the School of Public and International Affairs and is associated faculty at the Center for Information Technology Policy. Previously, she was an assistant professor at the University of Southern California, a research scientist at Google, and a privacy adviser at Snap, Inc. Her recognitions include a Presidential Early Career Award for Scientists and Engineers (PECASE), a Sloan Research Fellowship, and an NSF CAREER Award. Her work on RAPPOR-the first commercial deployment of differential privacy-received the CCS 2024 Test-of-Time Award. Her research on discrimination in ad delivery earned a 2019 CSCW Honorable Mention and Recognition of Contribution to Diversity and Inclusion, and it won the 2025 FAccT Best Paper Award. She holds a bachelor's degree from MIT and a Ph.D. from Stanford University.
What to watch next
The panel will convene over the next three years and release annual reports at the U.N. Global Dialogue on AI Governance. For researchers and R&D leaders, this is a clear signal: prioritize evaluation methods that travel well into policy, strengthen documentation and auditing pipelines, and make privacy and fairness guarantees explicit.
Build skills aligned with this agenda
If you're updating team capabilities in auditing, evaluation, privacy, or AI governance, explore curated learning paths and fresh releases:
Your membership also unlocks:
