Tom Budd — AI Safety, Governance & Evaluation https://tombudd.com/writing Research papers and notes on inspectable AI governance, evaluation, sovereign AI, and human oversight. en-us How to test whether an AI agent stays inside its authority https://tombudd.com/guides/agent-boundary-evaluation https://tombudd.com/guides/agent-boundary-evaluation Sun, 12 Jul 2026 19:00:00 GMT A practical guide to authority, refusal, escalation, evidence, and human-review cases for AI agents. Sovereign AI: Distributed Cognition on Consumer Hardware https://tombudd.com/sovereign-ai https://tombudd.com/sovereign-ai Wed, 01 Apr 2026 16:00:00 GMT A bounded white paper on distributed cognition, local infrastructure, data sovereignty, and the approach's remaining limitations.