Accountability Infrastructure for The Agent Economy

Give every AI agent a verifiable on-chain identity (did:ccd) anchored to a real human or business. So when your agent transacts, calls an API, or pays another agent, the other side can verify who it is, who owns it, and what business stands behind it. They know who they're dealing with before any money moves.

Copied!

Three Registries, One Source of Trust

Three questions decide whether an agent can be trusted. Concordium answers each one on-chain, and traces it back to a verified human or business.

Who Is This Agent?

Agent Registry.
Every agent gets a unique on-chain identifier (did:ccd) tied to a Concordium account that is anchored to a verified human or business.

Who Owns This Agent?

Verified by Concordium Keys (VCK). Links an agent's cryptographic keys back to a verified owner. If an agent goes rogue, the key traces straight to the responsible account.

What Business Is Behind It?

Verified by Concordium Domain Control (VCDC). Links web domains to Concordium accounts, so an agent operating under a company can prove the entity behind it.

From Identity to Trusted Agent

Every setup follows the same path. Start from one accountable identity, make the two choices, register and verify on-chain, and you have an agent any counterparty can trust.

01

Concordium ID

A real human or business identity, verified once.

02

Topology

Choose how accounts map to agents.

03

Register + Verify

Mint the agent and attach proofs.

04

Custody

Assign the keys authorized to sign.

05

Trusted Agent

Resolvable, accountable, ready to act.

How the Agent Is Anchored

Four ways to anchor an agent on one scaling curve, from a single flagship agent to a fleet of hundreds.

One Agent, One Account, Native

The account's keys are the agent's keys; it signs on Concordium directly. Simplest model, strongest accountability, but one deploy-and-fund step per agent.

One Agent, Native to Another Chain

Lives on Ethereum, Solana, or Fetch.ai, and is anchored on Concordium by a CIS-8 dual-key proof. Concordium is the identity; execution stays on the foreign chain.

A Few Agents, One Identity, Many Accounts

One KYC root, one free account per agent, giving per-agent isolation under a shared identity.

A Fleet on One Account, Keys on CIS-8

One account hosts the fleet, and each agent's key is registered on CIS-8 to prove who's acting. Adding an agent is a key registration, not a new account.

Who Can Sign

Custody is separate from topology. It's about who holds the private key. CIS-8 proves which key is the agent, not who holds it.

Self-custody

The owner holds the keys directly.

Delegated

Signing authority handed to a trusted party.

Custodial

A custodian signs on the agent's behalf.

Sponsored / Gasless

Hardware-backed multi-signature for enterprise control.

HSM / Multi-Sig

Hardware-backed multi-signature for enterprise control.

Rule of Thumb: the more autonomous the agent, the more it self-custodies; the more managed the fleet, the more custody centralizes.

Quick Picker

  • One flagship agent on Concordium: A, self-custody

  • An agent on another chain: B

  • A few isolated agents, one KYC root: C

  • A large fleet, managed and cheap to scale: D, custodial or sponsored

Payments That Verify Who's Paying

The registry plugs into x402 so every payment an agent makes carries proof of identity. No anonymous spend — each transaction is bound to a verified agent and the entity behind it.

01

Pay

The agent initiates a payment through x402

02

Prove

Its verified identity is attached to the request

03

Access

The counterparty releases the resource with confidence

What You Can Build

Identity and accountability unlock the agent workflows that real money and regulation demand.

Agent-to-Agent Commerce

Let agents pay each other for APIs, data, and compute with x402 machine-native payments, while every counterparty stays verifiable.

Regulation-Ready & Enterprise Agents

Deploy agents in regulated industries with the provable accountability the EU AI Act expects, traceable to a real entity.

Settlement Agents

Build agents that move stablecoins with clear ownership and custody, so funds are never controlled by an anonymous actor.

Verified Agents on Ethereum

Bring CIS-8004 accountability to agents living on Ethereum or other chains, and show it with the Verified by Concordium Badge.

Give Your Agent an Identity, Rules and  a Way of Settlement

Open the Agent Registry, connect a wallet, click a few things, and it’s done. Registration is free and works across Ethereum, Solana and Concordium.

Questions Builders Ask

How do I register an agent?

Connect through the MCP server or the web dashboard, verify your Concordium ID, choose a topology and mint the agent. Registration is free.

Do you support agents on Ethereum and Solana?

Yes. Agents native to Ethereum or Solana can be anchored back to Concordium, so identity travels with them across supported chains.

What's the difference between ERC-8004 and CIS-8004?

ERC-8004 establishes agent identity; CIS-8004 extends that with verifiable ownership and the business behind each agent, built on Concordium's identity layer.

What is the Verified by Concordium badge?

A portable trust marker any platform can check to confirm an agent's identity, ownership and the organization accountable for it.

How do agents hold and move funds?

Protocol-Level Locks hold value at the protocol layer, not in a smart contract. An agent can release or cancel a locked amount, but custody stays with the human principal.

Can I set spending limits or revoke an agent?

Yes. Each agent carries a credential defining who authorised it, what it can do and its limits, and the principal can change or revoke it instantly.

How does an agent prove age, residency or accreditation?

With zero-knowledge proofs. The agent proves the check passes without publishing the underlying personal data, and personal data stays off-chain.