RESEARCH
Research programmes
Our work is organised around a small number of focused programmes. Each turns a sharp question about AI-mediated power into something usable by regulators, institutions, and the public.
Admissibility & Evidence. Before asking whether an AI system is safe, we ask whether it should have been allowed to act at all — and on what record. This programme develops practical methods for evidence, traceability, and the conditions under which a system may enter real decisions.
Agentic Government & the State. AI is moving from advising to acting — running parts of the workflow in public administration, courts, services, and enforcement. This programme studies what happens to citizens, accountability, and the rule of law when the state’s decisions run partly through autonomous agents.
Access Classes & Routing Rights. People are rarely told “no” outright. They are simply not shown, not served, not prioritised, or quietly sorted by systems they never see. This programme maps these hidden “access classes” — and argues for a citizen’s right to know they were routed, and to be routed differently.
Futures (Foresight). Clearly labelled as foresight. Structured scenarios of how the relationship between human decision-making and AI could evolve — the paths worth preparing for, and the ones worth preventing.
Our outputs
Working papers, policy briefs, open diagnostic tools, and convenings. We publish openly and write to be read.
[ Read our latest insights ]
