Research and infrastructure for AI governance.
AI agent governance has a topology. We measure it across 543 regulated financial institutions. 90.8% sit inside a shared AI vendor mesh, bridged by 75 vendors. Credit ratings explain 1% of governance risk.
Published findings on 543 regulated financial institutions, 95,876 AI agents, 626,390 inter-agent edges. Four working papers covering the Coase Inversion, the credit-rating decoupling, systemic governance risk, and methodology validation. Cryptographically frozen substrate. Reproducible from a published methodology.
The deep stack underneath the research. Force-directed topology of the agent estate at scale. Semantic search across institutional governance posture. Persona-resolved intelligence and cascade modelling for top-tier engagements. Authenticated access only.