Trust Infrastructure for Digital Entities

The internet is entering a phase in which AI agents, not humans, increasingly decide what to read, buy, cite, and trust. No public infrastructure currently lets an agent verify who — or what — it is actually interacting with. TETA+PI is an open, cryptographically-verifiable trust registry for digital entities: people, businesses, brands, domains, APIs, AI models, MCP servers, and AI agents themselves. Critically, TETA+PI does not verify that an entity merely exists or is legally registered — it verifies specific claims an entity makes about itself, anchored to evidence captured at the moment of creation. Entities are verified against official registries and content-provenance standards (C2PA), with every verification event anchored to Bitcoin OpenTimestamps. On top of the registry, we introduce TWIRA (Trust-Weighted Intent Routing Algorithm), a ranking method that routes agent queries to verified entities using a combination of trust, semantic intent alignment, and provenance integrity. We describe the registry model, the claims-based verification architecture, TWIRA, and a structural property we call the Temporal Moat — an append-only chronology of verification events that cannot be reconstructed retroactively by a later entrant. This paper states plainly what is implemented today, what is proposed and awaiting evaluation, and what remains an open research question.

Read the full paper

10 pages. Registry model, claims-based verification, TWIRA, the Temporal Moat, roadmap, and governance.