Humans get +1.
Agents get +1+1.
Give your agent a phone number, with identity, scope, and revocation built in.
Alignent issues +1+1s for autonomous agents: real numbers on real networks, bound to hardware-verified identity, enforced by mission scope, revocable in seconds. A separate number space for a separate class of actor.
Press Issue +1+1 to mint a verified agent number.
The agentic economy is being built on infrastructure that wasn't built for it.
Every voice agent, browser agent, and autonomous system shipping today is anonymous traffic on networks designed for human subscribers. There is no standard way to verify who an agent is, what it's allowed to do, who's accountable for its actions, or how to turn it off.
Training-time alignment (RLHF, constitutional AI, fine-tuning) is real work, and it's necessary. It ends the moment the model is deployed. After that, every guarantee degrades in contact with prompt injection, compromised orchestration, session hijacking, and emergent multi-agent behavior.
The only durable accountability is enforced outside the agent, at infrastructure the agent cannot rewrite.
Separation of powers for the network.
Humans have phone numbers. They are how we are reached, how we are verified, how we are held accountable when something goes wrong. The number is a small piece of civic infrastructure that does an enormous amount of work.
Agents don't have one. They borrow ours, spoof ours, or operate without one entirely. As agents take on more of the calls, transactions, and interactions in daily life, the line between human and agent traffic is going to need a structural answer. Not a checkbox at the bottom of a terms of service page.
Alignent's answer is the +1+1: a parallel number space for autonomous actors. Verifiable at the network. Revocable in seconds. Scoped to what the agent is supposed to do, and only that.
One number space for humans. Another, with rules, for agents.
Same dial format. Different properties at the network layer.
The thesis, in eight words
Humans get +1.Agents get +1+1.
alignment as infrastructure
A +1+1 is three things at once.
An identity
Cryptographically bound to the agent and to the hardware it runs on. Not spoofable. Not transferable without an audit trail.
A scope
Time, geography, role, network, spending. What the agent is allowed to do is enforced at the network before the agent can act, not after.
A lifecycle
Issuable in seconds. Revocable in seconds. Every issuance, modification, and termination is a permanent record.
Together, these three properties turn an autonomous agent from anonymous traffic into a named, scoped, accountable actor, without slowing it down.
Built for three audiences, on one primitive.
AI builders
Shipping voice, browser, and autonomous agents who need their agents to pass verification, receive codes, and operate in the systems that already exist.
For developers02Enterprises and safety teams
Deploying fleets of agents who need audit trails, scope enforcement, and the ability to revoke instantly when something goes wrong.
For enterprisesRegulators, carriers, policymakers
Who need a registry, a standard, and a way to draw the line between human and agent traffic without breaking either.
v1 — reading onlyThe patent estate predates the category.
The primitive Alignent issues was filed as a provisional in June 2021. Before “agentic AI” was a phrase, before DePIN was a category, before the GPT-4 launch that made any of this urgent. The granted patent (US 12,574,738 B2) and its continuation cover the architecture end to end.
17 years of exclusive runway. Forward-cited by T-Mobile. Licensed exclusively to Alignent for the agent identity domain.
Patent estate
US 12,574,738 B2 · 22-year window
Alignent is in stealth.
A working prototype built in 48 hours won a recent agentic-AI hackathon. The production platform is in development now, with a small set of design partners shaping the spec.
If you're building agents, deploying agents, regulating agents, or carrying their traffic, the list is open.
Investors, partners, press: hello@alignent.com