A series on what it actually takes to build and run AI systems under European sovereignty constraints. Covers the strategic case for jurisdiction-aware architecture, the UK-EU regulatory split with compliance-as-code implementations, and the practical infrastructure stack for sovereign inference.

All Posts in Order

  1. Building AI Platforms Under European Sovereignty

    Why jurisdiction, not model capability, is the real constraint on European AI. How to architect AI platforms that sta...

  2. Architecting for Two Europes with Compliance-as-Code

    How to build one AI platform that satisfies both the EU AI Act and the UK's lighter regime, with the regulatory split...

  3. A Practical Guide to the European Sovereign Inference Stack

    Running LLM inference on European sovereign cloud providers instead of AWS, GCP or Azure. Who offers GPUs, what it co...

  4. Securing the Open Source AI Supply Chain

    Securing the open source AI supply chain. Model weights, pickle files, PyTorch dependencies and how to verify what yo...

  5. What It Actually Takes to Fine-Tune an LLM in Europe

    What it legally takes to fine-tune an LLM on proprietary data in the UK and EU. Copyright, text and data mining excep...

  6. When the API Gets Switched Off

    Anthropic disabled its newest models outside the US overnight. What that means for UK AI sovereignty, and why buildin...

  7. Compute Isn't the Constraint

    Why the UK's sovereign compute fund misreads the frontier. Research talent, not GPUs, is what separates labs that lea...

  8. Where AI Research Actually Happens, in Six Charts

    Publication share, top-cited papers, talent flows, private investment and the open-weight lag, as six charts with the...