ATAH · Agent to Authenticated Human Protocol
The open protocol for AI-to-human handoff.
When an AI system reaches the edge of what it should answer on its own — a legal question, a complex insurance risk, a tax decision — the journey usually stops at a disclaimer. ATAH is the open protocol for handing off to a verified human professional, with trust, consent, and lifecycle defined. It composes with MCP and the wider agentic stack — exposed natively as MCP tools, with a REST contract alongside.
Draft v0.8.2 (rc) · Apache-2.0 · open for review · reference implementation in active development
- version
- 0.8.2 (rc)
- licence
- Apache-2.0
- status
- release candidate
- governance
- public charter
The problem
The broken handoff
AI systems can now book flights, complete transactions, and run multi-step workflows on behalf of users. But when an AI identifies that human professional expertise is genuinely needed — a lawyer for a contract, a financial advisor for a regulated decision, a licensed agent or broker for a complex insurance risk, an engineer for a building issue — the journey breaks. The AI typically stops at a disclaimer and leaves the user with a search box.
Three sides to the same gap
- AI platforms want to complete the journey, not end it on a disclaimer.
- Users need a verified human at the moment they most need one — not a search box.
- Professionals need fair discoverability as AI replaces search as the first call. Without a neutral layer, only the largest commercial brands surface inside AI conversations.
That broken handoff is the gap ATAH is designed to fill.
The protocol
What ATAH defines
ATAH is an open protocol layer. It gives AI systems a structured, trusted way to connect users to credentialled and established human professionals, with verified data drawn from authoritative sources — regulators, professional bodies, independent verifiers — and provenance attached to every claim.
The protocol has two components:
Component 1 — Discovery. A query interface for AI agents: "find verified professionals meeting these criteria." Returns a candidate set with source-tagged provenance, verification status, and a structured non-recommendation disclosure. ATAH does not rank, score, or recommend; the AI platform decides which candidates best fit the user's situation, in what order to present them, and how to surface them — using ATAH's verified data and provenance as the basis for that judgement.
Component 2 — End-user handoff. The lifecycle for moving an end-user (an individual or a business) from a Discovery result to a working introduction with a chosen professional, with structured consent, vault-mediated PII flow, and crypto-erasure on retrieval.
Provenance always visible
Every claim about every professional is sourced and dated. AI platforms see the basis of trust; users see it too. Nothing is collapsed into an opaque score.
No commercial weighting
Hard filters for eligibility, transparent verification bands, randomisation within each band under a documented fairness policy. No paid placement, no commercial influence on ordering.
Open and neutral
Apache-2.0 specification. Public Charter with eight entrenched commitments. Designed to be operated by multiple registries over time.
How it works
The ATAH protocol flow
A single diagram of the two components in motion: an AI agent calls Discovery and receives a candidate set of verified professionals with provenance attached; optional handoff then proceeds with structured consent, a transient PII vault, authenticated retrieval, and crypto-erasure on completion.
The agentic stack
Every other agentic protocol is machine-to-machine. ATAH is the human one.
MCP gives agents tools and data. A2A lets agents talk to other agents. ACP and AP2 cover agent-mediated commerce and payments. OAuth and Verifiable Credentials handle identity and credentials. All machine-to-machine. ATAH is the missing layer for handing off from an AI agent to a verified human professional when that's what the situation actually needs.
ATAH composes with this stack rather than replacing it — it's exposed natively as MCP tools (and as REST), so AI platforms already speaking MCP call ATAH directly, with no separate data protocol to wire up.
Audiences
Built for
For AI platforms
Implement via MCP or REST. Surface verified professional candidates with structured provenance and non-recommendation disclosure on every response. Royalty-free.
Read more →For professional bodies
Make member or licensee status discoverable on neutral infrastructure with public governance. Public partner fee schedules; waivers for regulators and public-interest bodies.
Read more →For professionals
Two routes to representation — through your professional body, or by individual self-registration. Identical matching treatment between routes.
Read more →Where we are
Status
ATAH v0.8.2 (rc) is the first published version, released as a release candidate for technical review and reference implementation work. The full schemas, OpenAPI contract, MCP tool definitions, conformance documentation, and Charter are all on GitHub under Apache-2.0. Anyone may implement, fork, or extend.
Breaking changes are possible until v1.0 per the published version negotiation policy. Current version, full changelog, and roadmap are tracked in the repository.
Author
About the author
ATAH was authored by Grahame Cohen, founder of Epoq and CEO of Epoq North America. Epoq has spent more than twenty-five years building digital legal services for consumers and small businesses, working alongside major insurance, financial services, and other large partners across the UK, US, Canada, and other markets.
The protocol is published independently of Epoq under Apache-2.0 with a public Charter, and is designed to be operated by multiple registries over time.
Press · Speaking · Interviews
For journalists and conference organisers
ATAH's author is available for journalist interviews, conference speaking opportunities, and panel participation on AI-to-human handoff, commercial neutrality in AI-mediated discovery, and the implications for insurance, legal, financial, and other regulated and established professional sectors. Bio, headshot, quotable framings, and downloadable assets are on the press page.