Proof of Humanity promises to fix the web’s most persistent problems. One human, one account, one digital life. No more bot armies, no more fake engagement, no more scams.

And I should be the last person to doubt it. My team has spent years building the cryptographic tools that make these systems possible. We pioneered passport ZKs (a way to prove you hold a valid passport without revealing your personal data). We built the first zkML for biometrics (a technique where your face can be recognized by a machine-learning model, and you can prove without exposing the raw image). If anyone had the tools to make Proof of Humanity work, it is us.

And yet, I’ve come to believe it will almost certainly fail not because the math is broken, but because the assumptions are.

When people talk about PoH, they often gesture vaguely at “Sybil resistance” without clarifying what problem they’re actually trying to solve. But most use-cases fall into three categories: economic fairness, governance legitimacy, and social trust. It feels clean in theory, but the moment it hits the real world, it warps.

The idea is simple: tie one account to one human, and abuse disappears. But each barrier you add; a scan, a credential only raises the payoff for getting around it.

Take crypto airdrops. They were designed to reward real users, yet bot farms evolve weekly, outsmarting every new defence. Today, almost every campaign profile I check on X carries a blue check mark. In theory, that means accounts are tied to real bank cards. In practice, verification hasn’t ended abuse; it’s created a premium market for it.

Or look at gambling. Every player is already KYC’d. Yet in 2024, operators lost $2.8 billion to fraud, most of it from bonus abuse, not fake humans, but credentialed, verified humans gaming incentives at scale.

Why does this happen? Because PoH raises marginal costs instead of eliminating them. To work, systems like this need the cost of extra identities to fall to zero or even negative, so there’s no benefit to pretending to be two people instead of one. But PoH does the opposite: every credential becomes an investment with a potential return. Abuse doesn’t go away — it scales.

The problem isn’t bots. The problem is that whenever you build a gate, you also create an economy around breaking it and you’re stuck in a loop.

Another common claim: PoH is supposed to strengthen democracy and DAO governance by enforcing “one person, one vote.” If everyone is verified as unique, then collective decision-making is fair.

But history shows otherwise. Cleaned-up voter registries and credential checks were supposed to protect democracy. Instead, they just commercialized it in most parts of the world. It didn’t prevent manipulation; they simply put a price tag on it.

And there’s a deeper flaw here: “one person, one vote” is the most non-contextual rule in coordination. Real governance is always layered. Who gets to participate? Under what conditions? With what weight? In every system, legitimacy isn’t established by uniqueness alone.

If passports are used as trust anchors, uniqueness is naturally enforced through voter registries. You don’t need an additional universal PoH system layered on top , the governance logic itself enforces uniqueness within context.

The same pattern would play out in DAOs. A healthy system doesn’t start from a flat assumption of “one person, one vote.” It defines who counts as a person in this context and how their participation should matter. Proof of Humanity skips over that nuance and tries to universalize identity, which is not enough for legitimacy.

But the deepest flaw is that PoH is positioned to find a way to restore authenticity to online life. What most of us resent online isn’t the presence of “non-humans.” It’s the flood of contextless noise, low-effort comments, swarm campaigns. Even if every account were tied to a verified, unique individual, the quality of conversation wouldn’t magically improve. Because this isn’t an identifier problem at all, it’s an algorithmic one.

And just like in governance, the solution is context, not universality. Social interactions always depend on layers of relevance: who is speaking, to whom, about what, and under what conditions. A like, a reply, or a review only makes sense when filtered through that context. A universal PoH stamp doesn’t help because it flattens everyone into the same category of “verified human,” while what we actually need are filters tuned to intent, relationship, and trust.

That’s why the real answer looks less like a checkpoint and more like a filter. Email solved its spam crisis not by demanding passports or eye scans at the door, but by building adaptive filters that learned over time to spot junk with near-perfect accuracy. Today 99.9% of spam never hits your inbox.

That’s the future of social platforms too. With AI, filters can become personal: tuned to your values, your preferences, your thresholds. They can spot patterns, curating what reaches you without putting hoops in front of those you want to hear from. (that’s a separate blog)

But that vision is very different from the fortress mentality of Proof of Humanity.

Not the Matrix “Zion” strategy of hiding behind walls.
The Neo strategy: stepping into the chaos armed with better tools.

PoH feels like progress because it’s tangible, measurable, and easy to pitch. But it rhymes with centuries of failed attempts to “purify” participation in human systems, whether in democracy, in decision-making, or in access to rights.

Every one of those experiments failed, and often backfired, because universal filters are neither scalable nor fair. They are always gamed, always politicized, always abused. Proof of Humanity risks replaying that history, this time on a digital scale.

The future of the internet will be shaped by invisible systems that won’t ask you to prove you’re human — it will simply feel human again.