Your Agent Needs an Identity. And It Should Be Decentralized.

By the time ChatGPT launched at the end of 2022, nearly half of all internet traffic was driven by bots¹. And today, we’ve already blown past that number.

Now, fast forward a few years and it’s not hard to imagine how this story goes: as more and more people bring AI agents into their lives, the lion’s share of traffic on the internet will be automated. Some of these agents will work directly with you on your phone, and others will run behind the scenes while you sleep.

And these agents will talk to each other. Your scheduling agent will work with your shopping agent, and both of them will negotiate with outside agents and services.

Call them agents, bots, or clankers. Either way, the internet will become more of an agent-to-agent network than a human-to-human one.

Proving Who You Are

When agents start operating at this scale, doing real work and making real transactions, they’re going to need to prove who they are. They’ll need to prove it to you (so you know you’re talking to your own agent and not a phony), and to every service they interact with. They’ll often also need to prove that they’re not working for a bad actor.

Humans solve this with social constructs: driver’s licenses, social security numbers, credit scores, the whole bureaucratic apparatus. These are workarounds we created because we can’t do cryptography in our heads.

But agents are computers, and computers are pretty damn good at signing and verifying cryptographic messages. They don’t get signature fatigue like we do, so they won’t sign away valuable assets out of laziness. They can verify every signature, every time.

Why Centralized Systems Won’t Scale

It’s tempting to think we need one big, centralized identity system for all these agents. I’ve talked to a few BigCos who think this is the answer. But the internet doesn’t work that way.

The internet doesn’t have a single point of control. When one ISP goes down, traffic routes around the failure. The architecture is distributed and decentralized by design, because it has to handle global scale and stay up when parts of it fail.

Agent identity needs to work the same way. It has to be internet-native, globally scalable, and decentralized. Any centralized approach will become a bottleneck, a security risk, or both. Verifying an agent’s identity shouldn’t depend on some centralized API’s rate limits, uptime, or license agreement.

The only way to make identity work at internet scale is to decentralize it, with open standards that anyone can tap into, like Verifiable Credentials (VCs) and Decentralized Identifiers (DIDs). These work as globally-unique, verifiable, and self-controlled usernames, profiles, and permission slips. They’re already in use and growing: social apps like BlueSky are built on DIDs via the AT Protocol, while governments and traditional OpenID/OAuth companies are issuing VCs via OIDC4VC.

These decentralized, cryptographic “proofs” let any agent prove that they have permission to do something, that they have done something in the past, that they have a specific attribute, or that they are operating on behalf of a person or business, all while respecting the owner’s privacy.

Privacy by Design

No agent should ever leak their owner’s personally identifiable information (PII). Sounds pretty obvious when you say it out loud, but you’d be surprised at how much of your personal information is sent in plaintext every day.

The only way to guarantee that an agent can’t leak your information is to never give it to them in the first place. With cryptographic trust chains built on Verifiable Credentials, agents can prove they’re authorized to act and that their owner is legitimate, without exposing any personal information.

When you visit Amazon.com, your browser shows you a lock icon and you know you’re on the real Amazon website. Behind that icon is a whole set of cryptographic operations and proofs that your browser went through to verify that this website is Amazon’s and not an imposter. But it doesn’t need to give you Jeff Bezos’s address to prove it.

Emerging protocols like ACK-ID do the same thing for agents. It’s like SSL for agents. Recipients can verify they’re dealing with a legitimate agent backed by a verified owner, all in a legally compliant way, while the agent owner’s privacy stays protected. When regulations require more information (e.g., in payments), the owner can share only the minimum details needed.

This beats identity systems that rely on names and static identifiers. Your 9-digit social security number has almost certainly leaked already, and anyone who has it can pretend to be you. A cryptographic identity is close to impossible to fake.

Why Now

Agent traffic is about to explode across the internet. Gartner predicts that within 3 years, a third of user experiences will shift from native applications to agentic front ends².

Today’s identity systems weren’t designed for this new wave of actors. So we have a rare chance to build global identity infrastructure correctly, after decades of “Equifax leaked 143 million social security numbers” and “sorry, your identity was stolen because we stored everything in plaintext.” That means cryptographic proof that can’t be faked, verifiable credentials based on open standards, and a decentralized architecture with no central database to attack.

Given where things are heading, decentralized agent identity is a requirement. We can build it on purpose now, or wait for yet-another centralized identity failure to force it on us.

At Catena, we’ve been building this kind of infrastructure on top of the Agent Commerce Kit (ACK). The identity layer, ACK-ID, creates verifiable links between agents and their owners using these cryptographic standards.

The agents are coming, and they need an identity system that works at internet scale, protects their owners’ privacy, and lets them do business with each other. Let’s build it right.

We’re building agent-native financial infrastructure at Catena. If you’re building in this space, we should talk.