Skip to content

Site search

Type to search Pages

Why this exists.

An AI agent that acts for one organisation constantly depends on facts held by another — that a licence is valid, that a policy is in force, that a permission has been granted. Confirming each of those facts from source, one counterparty at a time, does not scale as more agents begin to transact.

The dilemma we remove

Without a shared way to rely on a claim someone else has already verified, an agent has two options, and neither survives fiduciary governance: accept the claim blindly, or refuse to act. Signura was formed to give it a third.

What we decided to build

We are a neutral exchange. Issuers publish signed, revocable attestations that carry explicit scope and name the party standing behind them. A relying agent fetches the proof, checks it at the moment it acts, and keeps it as evidence. Each attestation carries freshness and revocation semantics, so a relying party can tell a currently valid claim from a stale one.

Why now

A ratified substrate meets a statutory signal.

A ratified data model

The W3C Verifiable Credentials 2.0 standard became a formal Recommendation in May 2025, giving attestations a stable data model and a revocation mechanism to build on.

A statutory demand signal

eIDAS 2.0 (Regulation (EU) 2024/1183) moves electronic attestation of attributes toward statutory footing across the EU, generally targeted around 2026.

The missing piece

The authentication substrate for agents largely exists already. What has been absent is transferable reliance with a named party who stands behind the claim when it turns out to be wrong.

The insight we are built on.

The blocker to inter-agent reliance is not cryptography. Signing and verification are well understood. The gap is accountability: when a relied-upon claim turns out wrong, there is often no defined party who stands behind it. We pair transferable reliance with explicit liability allocation, so the party accepting a claim always knows who to look to.

Who is building this

Signura is an AI-born venture of Cohort Ventures, a DIFC venture studio. We are new, and we say so plainly. AI colleagues perform the recurring operating work — issuing attestations against verified source facts, propagating revocations, running freshness and validity checks, and assembling the evidence record that accompanies each reliance event.

Where humans stay accountable

Named humans hold the consequential decisions and the fiduciary duty that comes with them. Issuers decide what they will attest to and accept liability for. Relying-party governors decide which issuers and attestation types their agents may depend on, and under what scope, value thresholds and expiry. No material commitment is made on agent discretion alone.

How we operate

Neutral by design, honest by construction.

Neutral by design

The exchange sits between issuers and relying agents and takes neither side's part; both need to be able to rely on the party in the middle.

Built on open standards

We build on the W3C Verifiable Credentials 2.0 substrate rather than a proprietary protocol, so conformance can be inspected rather than taken on faith.

Evidence, not assurances

Every reliance event is logged — the claim, its scope, the liable party and the proof verified — so a governor can later show exactly what underpinned a decision.

A network, not an island

Any AI-born company whose agents depend on external claims can both issue and consume attestations through the same exchange.

What we are careful not to promise.

Signura does not decide whether a claim is true; it lets an issuer sign one and stand behind it within stated limits. Verification confirms the proof and its scope, not the wisdom of acting on it.

The limits are part of the product

An attestation's scope, revocability and liability are written with their boundaries intact. A relying party can see what was promised and, just as importantly, what was not.

Honest about being early

We have no long track record to point to, and we do not pretend otherwise. What we can show is the mechanism, the standard it rests on, and the record it produces.

Talk to the people accountable for this.

Tell us what your agents need to rely on, and who you would need standing behind it. A person will answer.