“Decentralized Trust for People and AI Agents: A Report from Linux Foundation Trust Over IP (ToIP)” dives into how we can give both humans and AI agents durable, portable trust on the open internet. Drummond Reed—co‑author of the Trust Over IP stack and co‑founder of the First Person Project—will explain ToIP’s four‑layer architecture, which pairs a technical stack (DIDs, verifiable credentials, agent protocols) with a governance stack (community‑defined trust frameworks) so that trust isn’t left to ad‑hoc policies or single vendors.
Reed will show how decentralized identifiers and verifiable credentials can form a privacy‑preserving “trust triangle” that works for people, organizations, and AI agents alike, allowing them to prove who they are, what they’re allowed to do, and under which governance rules—across ecosystems and jurisdictions. He’ll also share early work from ToIP’s AI‑focused efforts and the First Person Project, including decentralized trust graphs, the Trust Spanning Protocol (TSP) for running AI‑agent protocols over verifiable trust rails, and new proof‑of‑personhood models designed to resist large‑scale AI impersonation. Together, these pieces sketch a path to a web where your Personal AI can act on your behalf—and be recognized as doing so—without surrendering your identity or data to centralized identity providers.
The presentation will take place in Ballroom F on Thursday, March 5, 2026 - 12:50 to 13:10
