Toby Drinkall is a researcher at the University of Oxford’s Internet Institute and a Fellow at the Next Frontier Seminar, working at the intersection of computational social science, artificial intelligence, and national security. Toby develops evaluation frameworks to benchmark how large language models behave in high-stakes military decision-support simulations, with a focus on escalation dynamics and compliance with international humanitarian law.
Toby has contributed to UK Ministry of Defence–funded projects on AI in national security, directed AI strategy for a Washington, D.C.–based consulting firm, served as a panel specialist on LLM wargaming for the European Leadership Network, and presented his research at the Hoover Institution at Stanford University and other leading transatlantic policy forums.
Toby holds a First-Class degree with Distinction in Theology and Religion from Oxford and will be joining Palantir in 2025.