The handshake protocol the agentic web needs.
Bots overtook humans on the open web this month — faster than even the people closest to the metrics had forecast. The trust layer for what comes next can't be single-vendor. TruePE is convening a closed roundtable on the interoperable handshake protocol — and on what the steward-disposition operators across edge, identity, regulation, and category leadership need to do, together, to make it real.
Velocity is the story.
The forecasts kept compressing. End of 2027 became early 2027 became this month. Agentic traffic is now most of the open web. Every protocol assumption built on the premise of human traffic — identity, intent, attribution, rate-limits, consent — is now upside-down.
The honest position: nobody currently has the trust layer that the next phase needs. Not the edge providers. Not the identity-infrastructure operators. Not the regulators. Not the operating partners. Not TruePE. The shape of the answer has to be assembled across all of them — interoperable, cross-vendor, no single-provider moat — or the answer doesn't hold.
Working substrate, not press release.
TruePE's thesis is that integrity scales and manipulation collapses — and that the only way to prove either claim is to be them, not to write about them. The roundtable shape is the test of that proposition at the protocol layer. Working sessions among principals with skin in the answer. Chatham House by default. Outputs published only where the participants jointly sign off.
The convening is not a vendor pitch. TruePE is the steward of the convening. The protocol that emerges, if one does, belongs to the participants, not to TruePE.
Four questions, in this order.
1 · Interoperability without lock-in
How does a handshake protocol pass cross-provider and cross-vendor — Cloudflare, Akamai, Fastly, identity infra, sovereign endpoints — without any of them gaining a single-vendor moat that breaks the rest of the system? Past precedents: OAuth, WebAuthn / FIDO, C2PA / CAI. Failure modes: IAB Tech Lab fragmentation, every-vendor-their-own-SSO.
2 · Identity authentication, not surveillance
The agentic web needs identity that an agent can present and a counterpart can verify — without the protocol collapsing into a surveillance substrate. Hardware-bound credentials. Selective disclosure. Provenance trails. The lessons from WebAuthn and from C2PA both land here.
3 · Regulation that holds shape across jurisdictions
UK, EU, US, MENA, APAC. Each is moving on agentic AI on different timetables with different mechanisms. What does the regulatory substrate need to look like for an interoperable handshake protocol to hold across them, rather than fracturing into jurisdictional splinternets? What carrots and sticks already exist; what new instruments are needed.
4 · Cohort-level accountability and value devolution
The diagnosis everyone's converging on — that AI was built on collective data and the value should reflect that — has both extractive and structural answers. The protocol layer the roundtable is convened on has to make the structural answer technically possible: cohort-level attribution, audit-trail-clean provenance, voluntary devolution mechanisms that don't require forced equity transfer.
Categories, not nameplates.
Edge providers
Cloudflare, Akamai, Fastly, the smaller specialist edges. The protocol either passes cross-provider or it doesn't hold.
Identity infrastructure
FIDO Alliance principals, passkey platform operators, sovereign-ID providers, the privacy-preserving credential research community.
Regulators + policy
UK ICO, ICO-equivalents in the EU + US + MENA, the standards bodies (W3C, IETF, ISO TC 307, OECD AI Policy Observatory), select select-committee staff.
Operating partners
The carriers, retailers, energy operators, and banks who are about to deploy agentic AI at sovereign and regulated scale and need the trust layer to be real.
Steward-disposition principals
Operators across academia, the press, sovereign capital, philanthropy, and policy-shape who are holding the long-arc question. Carrier disposition matters as much as commercial position.
Three sessions across late 2026.
The convening runs as three working sessions across late 2026 — UK-anchored, hybrid-accessible for the principals who can't travel, Chatham House by default. Outputs are co-signed working documents, published only where participants jointly agree.
Invitations are extended on a steward-cultural-validation basis — the substantive part of the test is not credentials, it's carrier disposition. The roundtable will be small. Numbers are not the point; the right principals across the right categories are.
Registration on this page is the interest-capture step. Invitations are extended directly by Ian Daniels as Founding Steward.
First-party, passkey-authenticated.
The integrity protocols TruePE asserts at the fund layer apply at the registration layer too. Your interest record lives in our first-party store, accessible only to Ian. Authentication is hardware-bound via passkey — no shared secret in transit, no third-party identity broker. The friction is the filter.
This page sits at Level L1 (public) on the TruePE promotion ladder. Interest registrations arrive at L2 (warm network) and promote to L3 (counterparty under Chatham House confidentiality) on invitation issuance. The convening operates under the TruePE Mycelium sovereignty boundary — third-party-LLM exclusion applies to all convening materials; the published outputs are co-signed by participants under jointly-agreed scope.