a16z recently published an article arguing that AI agents need blockchain infrastructure to function in the real world. The argument is straightforward: as AI gets better at impersonation, we need cryptographic proof that an agent is authorized, that a payment is legitimate, that content is authentic.
I read it and thought: we've been solving the wrong problem first.
We're building cryptographic passports for bots. Meanwhile, humans still prove their professional skills with a PDF and a LinkedIn headline.
The problem a16z describes
The core argument goes like this. AI agents can now fake voices, faces, writing styles, and entire social personas. Traditional verification breaks down because AI improves faster than the tests designed to catch it. So you need infrastructure that makes impersonation expensive, not just detectable.
Their proposed solution: decentralized identities on blockchain. An agent carries a portable credential — a set of cryptographic signatures that prove who it represents, what it's allowed to do, and how it can pay. These credentials can be verified by anyone, revoked by the issuer, and they work across platforms without a central gatekeeper.
It's a compelling framework. But here's what struck me.
People have the same problem. And it's worse.
When an AI agent misrepresents itself, the damage is a bad API call or a fraudulent transaction. When a person misrepresents their professional identity, the consequences ripple through organizations for months or years.
Yahoo's CEO Scott Thompson resigned in 2012 after it came out he'd listed a computer science degree he never earned. He'd held the role for four months. Imagine the cost of that — not just the executive search, the board crisis, and the stock hit, but the decisions made during those four months by someone who wasn't who they claimed to be.
This isn't rare. Background check firms report that roughly 40% of resumes contain some form of misrepresentation. Not all of it is as dramatic as a fake degree. Sometimes it's inflated titles, fabricated project scope, or companies that no longer exist to verify against.
And we're not just talking about fraud. Even honest professionals struggle to prove what they actually know. The resume format forces a compression that loses most of the signal. You list "product management" but can't prove which decisions you made, which launches you led, or what your colleagues thought of your work.
The hiring side compensates with process: 5-8 interviews, reference calls, take-home assignments. Companies spend weeks reconstructing what a verifiable credential could confirm in seconds.
What a16z gets right, and where it applies to people
The a16z article identifies five things AI agents need from crypto infrastructure. Every single one applies to professional identity.
Raising the cost of impersonation. For AI agents, this means proof-of-personhood systems. For people, it means KYC-verified identity tied to professional claims. Not "I say I worked at Google" but "here's a cryptographic attestation, verifiable on-chain, that confirms it."
Decentralized proof of identity. The article warns that centralized identity systems become points of failure — a single platform can revoke access, impose fees, or enable surveillance. LinkedIn owns your professional identity today. If they change their algorithm, restrict API access, or shut down tomorrow, your professional history goes with them. A blockchain-based credential lives with you, not the platform.
Portable credentials across contexts. Professionals need to carry their authorization across chat apps, email, and APIs. Your proof of skills should work on LinkedIn, on a job board, in an email to a recruiter, and on your personal site. One credential, verifiable everywhere. AI agents need the same thing.
Programmable verification. For people, it means endorsements and skill verifications that carry weight because they're tied to real, verified identities. Not a LinkedIn recommendation that anyone can write for anyone — but a signed attestation from a verified colleague that specifies what you worked on together and what they observed. For agents, this means smart contracts that enforce payment rules.
Privacy-preserving proof. The article highlights zero-knowledge proofs — the ability to prove a fact without revealing the underlying data. For professionals, this is transformative. You could prove you earned above a certain salary threshold without revealing the exact number. You could prove you held a senior role at a Fortune 500 company without exposing which one, if confidentiality matters.
How this actually works in practice
The mechanics are simpler than they sound.
You verify your identity once through a KYC process — the same kind banks and crypto exchanges already use. Government ID, liveness check, done. This creates your master credential: a cryptographic key pair that lives on a blockchain and is permanently linked to your verified identity.
From this master credential, you generate sub-signatures for different purposes. One for your employment history. One for your skill endorsements. One for content you've authored. Each can be independently verified by anyone, traces back to your master identity, but doesn't expose more information than necessary.
When a former colleague endorses your skills, they do it with their own verified credential. The endorsement itself becomes a signed, timestamped, on-chain record. Not "John says Mike is good at product management." Instead: "Verified identity #4721 (resolved to John Smith, CTO at Company X) attests that verified identity #1193 (resolved to Mike Daykhin) demonstrated expertise in product strategy during their overlapping tenure at Company X from 2019-2022."
That's a fundamentally different kind of proof. You can't fake it. You can't inflate it. And you can verify it without calling John.
What this changes for hiring
Think about what happens when every professional claim is verifiable.
A recruiter looks at a candidate. Instead of scanning a resume and guessing, they see a verified timeline: confirmed roles at confirmed companies, with confirmed durations. Skills endorsed by people whose own identities are verified. Content authorship proven cryptographically.
The 5-round interview doesn't disappear entirely. Culture fit, communication style, specific problem-solving — those still need conversation. But the baseline verification? The "did they actually work there, do they actually know this, did their colleagues actually respect their work" questions? Those are answered before the first call.
For candidates, the benefit is just as real. You build your verified professional identity once. It travels with you. No more rewriting your resume for every application. No more "let me find three references who'll pick up the phone." Your proof of work exists independently of any platform, any employer, any recruiter's willingness to check.
The infrastructure isn't hypothetical
This isn't a thought experiment. The building blocks exist today. Blockchain identity systems, zero-knowledge proof libraries, verifiable credential standards (W3C has had a spec for this since 2019). What's been missing is the application layer that connects these tools to the actual workflow of hiring and professional reputation.
The a16z article ends with a statement: blockchains are the missing layer that enables trust at scale. They're mostly about that for AI agents.
But the irony is that people have needed this layer for much longer and I've been thinking, talking about and building around that almost 4 years now.
This article was inspired by the a16z crypto editorial "AI Needs Crypto — Especially Now." If you're interested in how decentralized identity applies to professional reputation and hiring, follow me and hit me a DM.