A Rare Opportunity to Shape Global AI Governance
In collaboration with Yoshua Bengio and MILA
Join a special round of the FIG Fellowship, built around three high‑impact research projects co‑led by Yoshua Bengio, Turing Award–winning computer scientist and one of the world’s most influential voices on AI safety and governance.
Fellows selected will work directly with Yoshua Bengio and senior policy leadership at MILA (Quebec AI Institute) on research to influence the trajectory of frontier AI.
This is a rare chance to contribute to the global conversation on AI safety—while co‑authoring serious research with one of the field’s defining figures.
Fellowship Projects
Insurance and Liability as Levers for AI Safety
How can financial and legal mechanisms drive safer AI development?
This project explores how insurance and liability regimes could be used to incentivise safety in the development and deployment of frontier AI systems. Drawing on lessons from sectors such as aviation, finance, and cybersecurity, the research will examine how risk pricing, liability allocation, and insurance requirements have historically improved safety outcomes—and how similar mechanisms could be adapted for AI.
The project will culminate in concrete policy recommendations for governments and regulators seeking to align incentives around AI safety without stifling innovation.
Output: A 10–20 page academic‑style report, co‑authored with Yoshua Bengio.
Ideal Candidates
Background in insurance, actuarial science, liability law, or cybersecurity risk
Experience in analysing how insurance or liability frameworks shape industry behaviour
AI expertise is a plus, but not required—we strongly encourage applications from insurance and risk professionals
❖
AI‑Driven Concentration of Power and Economic Sovereignty
What happens when frontier AI is controlled by only a few countries?
This project investigates how the concentration of advanced AI capabilities—primarily in the US and China—may erode the economic sovereignty of other nations. The research will analyse labour market disruption, economic dependency, and power asymmetries faced by countries without domestic frontier AI capacity.
Building on Yoshua Bengio’s existing work on international AI governance, the project aims to develop a rigorous economic framework for understanding AI‑driven global inequality and its policy implications.
Output: An academic‑style report (20+ pages), co‑authored with Yoshua Bengio.
Ideal Candidates
Background in economics, international political economy, or development economics
Experience with labour market analysis or technology‑driven economic transitions
AI knowledge is helpful but not essential—we are seeking strong economists entering a new domain
❖
Military AI and Threats to AI Safety
How do military uses of AI reshape global AI safety risks?
This project maps the rapidly evolving landscape of military AI applications—including autonomous weapons, battlefield intelligence, and decision‑support systems—and examines their implications for broader AI safety. The analysis will combine theoretical AI safety considerations with real‑world operational insights, including lessons drawn from contemporary conflicts such as Ukraine.
The goal is to better understand how military incentives, deployment environments, and escalation dynamics interact with long‑term AI safety concerns.
Output: A 10–20 page academic‑style report, co‑authored with Yoshua Bengio.
Ideal Candidates
Background in defence policy, security studies, military technology, or international humanitarian law
Familiarity with AI systems and AI safety debates
Experience with military or security applications of emerging technologies is a strong plus
About the Fellowship round
Format: Part‑time, remote‑first
Duration: 12 weeks
Time Commitment: ~8–10 hours per week
Outputs: Academic‑style policy reports, co‑authored with Yoshua Bengio
Audience: Policymakers, international organisations, and AI governance leaders
Fellows will be supported by FIG and Mila through structured supervision, research sprints, collaborative sessions, and direct engagement with project leads.