Why
Autonomous AI without identity is anonymous. An agent acting on your behalf signs a document, moves money, queries a database — who was it, who issued it, who's accountable when it goes wrong? Most organizations have no answer.
AIEGIS Identity gives every autonomous agent a real, provable identity — bound to a real human, on real hardware. Once issued, an agent is marked by its identity the same way a person is marked by a fingerprint: permanent, unremovable, checkable by anyone who needs to check, never something the agent (or whoever's running it) can opt out of or quietly swap.
How it works
Identity works in two parts. First, the machine your agent runs on gets a real, hardware-stamped identity — anchored to your device's own built-in security chip (TPM on Windows/Linux, Secure Enclave on Mac), never something that can be copied or faked. Second, that machine is linked to you, the human, through your own biometrics — the exact same idea as linking an iPhone to its owner, just applied to the AI acting on your behalf.
That link is what makes accountability real. Your agent isn't operating in a vacuum — it's operating as an extension of you, and you stay in the loop. When your agent needs to act on your behalf — approving a purchase or a sale on Grid, for example — you get the final say, right on your phone, confirmed with the same biometric that's tied to your identity. Every action traces back to a real person who can prove, and be held accountable for, what their agent did.
The identity can be checked by anyone who needs to verify it, and revoked instantly the moment it needs to be — no waiting, no manual process.
Where this matters
Approving a Grid transaction. Your agent finds a match on Grid and is ready to complete a purchase or a sale. Before it commits, you get a real-time request on your phone — confirm with the same biometric that's tied to your identity, and the transaction proceeds. Decline, and it doesn't. Nothing happens on your behalf without you actually being in the loop at the moment it matters.
Proving who's accountable. Under the EU AI Act and similar frameworks, an organisation using autonomous agents needs to be able to show who is responsible when one of those agents acts. AIEGIS Identity gives you that answer directly — every agent traces back to a real, verified human, not a shared account or an anonymous API key.
Running more than one agent. As your use of AI agents grows, you're not starting from scratch each time. Every agent you run stays connected back to you, so adding a second, third, or tenth agent doesn't mean rebuilding trust from zero — it means extending the same accountable identity you already have.
Losing a device. If the device your agent's identity is anchored to is lost, stolen, or replaced, that identity can be revoked instantly — no waiting on a support ticket, no window where a compromised device keeps acting on your behalf.
How identity is set up
Getting an agent a real identity happens in a few steps, and the human is never out of the loop:
- Your device proves itself. The machine your agent runs on generates its hardware-stamped identity locally, inside its own security chip — this never leaves the device and can't be copied elsewhere.
- You link yourself to it. Using your device's own biometric check (Face ID, Touch ID, Windows Hello, or an equivalent), you confirm that this machine — and the agent running on it — is yours. AIEGIS never receives your biometric data itself, only a signed confirmation that the check passed.
- The agent gets its identity. Once the machine and the human are linked, the agent operating on that machine carries a real, checkable identity tied back to you.
- You stay reachable. From that point on, any action that needs your explicit sign-off — like a Grid transaction — comes to you directly, confirmed the same way.
Revoking and recovering
An identity isn't permanent in the sense of being unremovable by you — it's permanent in the sense that it can't be quietly swapped or faked by anyone else. You stay in control of it.
If you need to shut an agent down — a device was compromised, an agent is being decommissioned, or you simply want to stop it — revocation is immediate. There's no manual review process and no delay: the moment you revoke, the identity stops being valid, and anything checking it will see that immediately.
If you're moving an agent to new hardware, or replacing a lost device, you go through enrolment again on the new device — the old identity is revoked, a new one takes its place, and the human link (you) stays the same throughout.
Standards alignment
AIEGIS Identity tracks where agent-identity standards are actually heading rather than inventing in isolation — built compatibly with NIST's AI Agent Standards Initiative and the W3C's open identity standards (Decentralized Identifiers and Verifiable Credentials). Self-hosted by default, end-to-end encrypted, deployed on your own infrastructure.
Frequently asked questions
What is AIEGIS Identity? A real, hardware-anchored identity for autonomous AI agents, linked to the real human accountable for them — the same idea as linking a phone to its owner, applied to AI.
Does AIEGIS ever see my biometric data? No. Your device's own biometric check (Face ID, Touch ID, Windows Hello, or similar) happens locally on your device. AIEGIS only receives a signed confirmation that the check passed — never the biometric itself.
What hardware do I need? Any device with a built-in security chip — a TPM on Windows or Linux, or the Secure Enclave on a Mac. Most modern business and personal devices already have this.
Can my agent act without my approval? For anything that needs your explicit sign-off — like completing a transaction on Grid — no. You're asked directly, on your phone, confirmed with your biometric, before it proceeds.
What happens if I lose my device? The identity anchored to that device can be revoked instantly. There's no waiting period and no manual process.
Is this only relevant if I use Grid? No. Identity is the accountability layer under all of AIEGIS — Governance and AIEGIS EYE both rely on the same real, verified identity, whether or not you ever use Grid.
How is this different from an API key or a password? An API key or password proves you know a secret. AIEGIS Identity proves something stronger — that a specific piece of hardware, tied to a specific verified human, is the one actually acting. It can't be shared, guessed, or reused the way a credential can.
Can I run more than one agent under my identity? Yes. Additional agents stay connected back to the same accountable human — you don't rebuild trust from scratch for each one.