Haydn Belfield

Haydn Belfield is a Senior Research Scientist (Frontier Planning) at Google DeepMind, working on Allan Dafoe's Frontier Safety and Governance Team. He joined DeepMind in March 2025 after eight years at the University of Cambridge, where he served as Research Associate and Academic Project Manager at the Centre for the Study of Existential Risk (CSER) and Co-Chair of the Global Politics of AI Project at the Leverhulme Centre for the Future of Intelligence. He is a DPhil candidate in International Relations at the University of Oxford, where his thesis — "Bombs, Bugs, Bytes and Bots" — examines what the varied success of arms control regimes for nuclear, biological, and cyber weapons can teach us about the international governance of artificial intelligence. He holds a BA in Politics, Philosophy and Economics and an MSc in Politics Research, both from Oxford. His earlier career included roles as Senior Parliamentary Researcher to a Labour Shadow Cabinet MP, Policy Associate at the Global Priorities Project, and the first Development Director of the Centre for Effective Altruism. He has over 40 publications with more than 2,100 citations, with key works including "The malicious use of AI: Forecasting, prevention, and mitigation" and "Toward trustworthy AI development: mechanisms for supporting verifiable claims," and has advised the UK, US, and Singaporean governments as well as the EU, UN, and OECD.

No endorsements yet.

Personal Website

https://www.haydnbelfield.com/

LinkedIn

LinkedIn

Twitter / X

Twitter

LessWrong

-

EA Forum

-