This project explores how AI risk scenario building can be made fit-for-purpose for government actors. The motivation is to help regulators and ministries develop anticipatory capacity, reducing the delay between rapid technological change and policy response. Rather than developing novel methodology, the project adapts existing frameworks, including national risk registers and reasonable worst case scenarios, drawing on AI incident repositories, capability and diffusion forecasts, and best practices from other high risk domains such as public health, or finance.

Alumni

Meet the authors

(Research Team Lead)

Isaak Mengesha

Isaak is a Postdoctoral Research Associate at the University of Oxford working on forecasting technological change.

His research examines how technological change interacts with institutional incentives and coordination constraints, shaping long-run welfare and development trajectories.

Drawing on agent-based modelling, network analysis, and empirical growth research, he focuses on decision relevant analysis for risk management, and AI governance.

Charlie Collins

A UK diplomat and civil servant of 10 years, Charlie emigrated to New York after a posting to the UK Mission to the UN.
Charlie completed the Taskforce Winter 2026 cohort (his team’s work on incident tracking was accepted into the 2026 ICML conference) and has continued to the Summer 2026 cohort, working to map government capacities needed to respond to plausible, severe short-to-medium-term AI scenarios.
Educated as an engineer, Charlie enjoys meaningful challenges that combine technical and policy problems, and working to solve these with smart, kind people.

Juan Felipe Cerón Uribe

Juan is an AI safety researcher at OpenAI, where he was a core contributor to the development and training of the first ChatGPT release.

He has co-authored frontier AI safety papers on adversarial robustness and scalable oversight, and represented OpenAI as an industry reviewer for the 2026 International AI Safety Report.

He is an expert in frontier AI development and red-teaming, and wants to contribute to AI governance.

Salvatore d'Ambrosio

Salvatore is an Italian-British strategy advisor with nearly a decade of experience guiding global corporations and public institutions, including the European Parliament, on foresight, sustainability, and innovation, informing decisions at the highest level.

His work now examines how AI can support and enhance strategic decision-making and the governance of AI risks from geopolitical and social angles, translating complexity into clarity and actionability.

Vickie Ellis

Vickie is a national security and evidence-driven policy expert who translates emerging AI risk research into government-ready frameworks and policy engagement strategies.

She is a Research Associate with Arcadia Impact's AI Governance Taskforce, researching AI risk scenarios and mapping the AI safety stakeholder landscape.

Previously, she spent over a decade at the Department of State managing $300M+ in U.S. investment and building the interagency relationships needed to get risk frameworks adopted.

Alumni

Meet the authors

Isaak Mengesha

(Research Team Lead)

Isaak is a Postdoctoral Research Associate at the University of Oxford working on forecasting technological change.

His research examines how technological change interacts with institutional incentives and coordination constraints, shaping long-run welfare and development trajectories.

Drawing on agent-based modelling, network analysis, and empirical growth research, he focuses on decision relevant analysis for risk management, and AI governance.

Charlie Collins

Juan Felipe Cerón Uribe

Juan is an AI safety researcher at OpenAI, where he was a core contributor to the development and training of the first ChatGPT release.

He has co-authored frontier AI safety papers on adversarial robustness and scalable oversight, and represented OpenAI as an industry reviewer for the 2026 International AI Safety Report.

He is an expert in frontier AI development and red-teaming, and wants to contribute to AI governance.

Salvatore d'Ambrosio

Vickie Ellis

Vickie is a national security and evidence-driven policy expert who translates emerging AI risk research into government-ready frameworks and policy engagement strategies.

She is a Research Associate with Arcadia Impact's AI Governance Taskforce, researching AI risk scenarios and mapping the AI safety stakeholder landscape.

Previously, she spent over a decade at the Department of State managing $300M+ in U.S. investment and building the interagency relationships needed to get risk frameworks adopted.

Programme

AI Governance Taskforce

The AI Governance Taskforce is a career development programme for experienced professionals looking to transition careers into AI governance, focussed on reducing risks from advanced AI.
Participants work around existing commitments during our 12 week, remote, part-time cohorts, producing policy research in teams of 4, led by our Research Team Lead staff in partnership with recognised experts in the field. Teams write an academic-style paper and accompanying blog post to build knowledge, skills and work portfolios.

Learn more