2026 Faculty Award Winners for Research at the Intersection of Technology and Geopolitics

The Belfer Center for Science and International Affairs at the Harvard Kennedy School of Government (HKS) and the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) has announced the second round of awards from their Faculty Research Fund on technology and geopolitics. 

Following a highly competitive process, with evaluation by a multi-disciplinary faculty committee, three proposals have been selected for funding. Awards have been made to teams led by:

• Arial Procaccia (SEAS), with Sharad Goel (HKS), and Jake Sullivan (HKS)
• Wolfram Schlenker (HKS), with Peter Huybers (SEAS)
• Juncheng Yang (SEAS), with Anthony Saich (HKS), Ryan Badman, and Jean Dong

The Faculty Research Fund—administered by the Belfer Center’s Emerging Technology, Scientific Advancement, and Global Policy program—provides support to researchers working collaboratively in areas where emerging technologies and scientific advances intersect with global policy and geopolitics. The selected projects are expected to produce publications, policy frameworks, curricular content, and novel AI-based tools.

This fund is part of the Program’s response to an urgent need to advance scholarship, build a new generation of leaders fluent in both public policy and technology, and test ways in which emerging technologies can be helpful in the processes of policy and geopolitics.

About the Projects

Project: Introducing AI and Generative Negotiation

Principal Investigator: Arial Procaccia, Alfred and Rebecca Lin Professor of Computer Science, Harvard John A. Paulson School Of Engineering And Applied Sciences

This proposal introduces “generative negotiation,” a framework for integrating generative AI into bargaining processes by allowing the AI not just to evaluate options but to expand the set of possible agreements itself. The project formalizes how AI can algorithmically generate new issues and policy linkages, thereby enlarging the negotiation space and potentially increasing joint gains. It combines theory with empirical evaluation and controlled geopolitical simulations. Its relevance lies in reframing negotiation as a computational search problem shaped by AI, with implications for institutional design, public policy, and international diplomacy, particularly in complex, high-stakes domains like technology governance and climate cooperation.  

Project: Exploring an AI-Based Early Warning System for Global Food Crises

Principal Investigator: Wolfram Schlenker, Ray A. Goldberg Professor of the Global Food System, Harvard Kennedy School

The disruption of reporting by the USAID-supported Famine Early Warning Systems Network (FEWS NET) in early 2025 highlighted a vulnerability in global food security infrastructure. We will examine the feasibility of an AI-based early warning system for global food crises that automatically pulls in freely available satellite, weather, and soil-moisture data to forecast harvests around the world and trace how shortfalls move through international trade and prices to affect import-dependent countries. This automated model will be tested against FEWS NET's historical expert forecasts from 2009–2020 to examine where data-driven systems can complement—or substitute for—expert judgment, and where they cannot.

Project: Integrated Technical-Policy Assessment of U.S. and Chinese AI Models

Principal Investigator: Juncheng Yang, Assistant Professor, Harvard John A. Paulson School of Engineering and Applied Sciences

AI leadership will depend not only on building the strongest frontier models, but also on achieving broad global adoption. In the past year, China has gained substantial momentum through cheaper, modifiable open-weight models. This project will undertake an integrated technical-policy assessment of U.S. and Chinese models, alongside surveys with practitioners in the Global South, to identify the factors beyond raw intelligence that shape choices among models and serving stacks. It will also develop a lock-in index to quantify long-term dependency across technical costs, hardware dependence, and country-specific factors.