Research
Research.
Policy briefs, papers and notes on sovereign AI compute. Every piece has a named lead and a target date.
Type
Topic
Published
Research plan
Dates are targets and may move.
Dec 2026
PlannedSiting sovereign compute: energy, grid and planningWhere new and existing compute can realistically be sited in the UK, given grid capacity, energy costs and planning constraints.
Infrastructure · Policy brief
Lead: Victoria de Bruijn
Jan 2027
PlannedAdapting open-weight models on British dataWhat it takes to adapt openly licensed models on British data, what they can be trusted with, and how they should be evaluated.
Models · Paper
Lead: Toby Chen
Jan 2027
PlannedTraining across ordinary networks: decentralised compute in practiceEvidence on training and serving models across many smaller nodes over standard connections, and what it means for UK capacity.
Compute · Paper
Lead: Emma Walker
Feb 2027
PlannedUsing the capacity the UK already hasAn assessment of compute already sitting in UK universities, public bodies and businesses, and how it could be pooled.
Infrastructure · Paper
Lead: Victoria de Bruijn
Mar 2027
PlannedAccrediting and auditing contributing computeA proposed accreditation scheme for organisations contributing compute, with an audit trail and the ability to roll back.
Governance · Policy brief
Lead: Emma Walker
Mar 2027
In progressRight-sizing models: routing requests by complexityHow sending each request to the smallest model that can answer it well reduces cost and energy without reducing quality.
Models · Note
Lead: Toby Chen
Apr 2027
PlannedProcuring sovereign AI in the public sectorPractical guidance for public bodies on specifying, buying and assuring AI services that run within UK jurisdiction.
Governance · Policy brief
Lead: Toby Chen, Victoria de Bruijn and Emma Walker