Cultural Essay

The Verified-Human Badge: what it means when a meeting asks you to prove you are a person

A growing number of everyday places — a video call, a hobby forum, a comment thread, a campaign ad — now ask the same quiet question: is anyone actually there? This essay follows that question from a Zoom meeting to an iris scanner, and asks what is lost and gained when being human stops being assumed and starts being shown.

By Culture desk × Folka

1. The room that checks your eyes

The meeting starts the way meetings always start. Someone joins, a name appears, a face settles into its small rectangle. Then a badge appears beside it: Verified Human. Nothing about the call looks different. But somewhere behind that badge sits a decision that would have been unthinkable a few years ago — that the easiest way to reassure a room is to confirm, cryptographically, that the person in it has a body.

None of this is paranoia. The fraud it answers really happened, and it worked because the call looked and sounded exactly like the real thing. But the fix is a strange kind of answer. It doesn't make the call harder to fake. It makes the people in it easier to certify — which is a different problem solved by a different means, and it changes what a meeting is asking of the people who join it.

2. What counts as proof now

For almost the whole of human history, being present in a room was its own proof. You did not need a badge to show you were a person; the fact that a body was standing there did the showing for you. Online, that assumption survived for decades on cues that were good enough — a voice that hesitated in the right places, a reply that arrived a beat late, a face that moved the way faces move. Those cues are no longer reliable, because the thing that can fake a voice can now fake the hesitation too. So the badge has quietly moved to the body: not what you can do, but what you are, checked against a record made in advance.

That question does not stay in the meeting room. It turns up wherever people used to be taken on trust — a hobby community, a comment section, a campaign ad — and each time it turns up, something different gets asked to prove itself.

3. The same question, everywhere

The pattern is not confined to video calls. It shows up in a decompilation hobby, on a social platform's data pipes, and in the middle of a political campaign — four unrelated places arriving at the same demand within months of each other.

A meeting, a hobby, a platform, an election: different stakes, same move. Something that used to be assumed human now has to be marked, and the marking is being built by the same handful of companies that also profit from selling the mark.

4. What fraud took, what recognition lost

It is tempting to read all of this as one story about fraud. It is not. The Zoom badge answers a financial crime — a fake executive, a real $25 million gone. The decomp.dev closure answers something quieter, and in its way sadder: not theft, but the loss of the ability to tell who had actually done the work.

decomp.dev's admin did not ban AI. He said plainly that the community aspect of decompilation had been broken — and that AI is useful when paired with people. His problem was never the tool. It was that he could no longer see, from the outside, who had done the seeing.

5. From passing a test to showing a body

This is not the first time the internet has had to prove someone was human. The CAPTCHA, built at Carnegie Mellon around 2003, asked you to read distorted letters or pick out a crosswalk — a small test built on the fact that a machine, at the time, could not do it either. For almost two decades, that was enough: proof of personhood was proof of a skill a machine lacked.

The CAPTCHA tested what you could do. World ID, and the badge built on it, tests what you are. That is the actual shift inside this story: from a performance a machine could not copy, to a credential a machine cannot hold at all — because a credential, unlike a performance, has to be issued, and issuing is something only an authority can do.

6. Mark the machine, or mark the person

Once the skill test stops working, there are really only two places left to apply pressure: the thing that was made, or the person who might have made it. Reddit is marking the platform's pipes. Zoom is marking the person. Those are not the same choice, and they do not cost the same thing.

Each path has a believer and a buyer. World sells the iris scan as proof. Reddit sells licensed, verified-human data to the same AI industry it is restricting. The pressure that created the doubt is also the market now selling the cure for it — a fact worth stating plainly, without assuming anyone intended it that way.

7. The case against certifying people

The strongest objection to certifying people rather than content is not philosophical. It is regulatory, and it has already happened. Thailand's data regulator ordered World to delete 1.2 million iris scans in November 2025. Kenya, Indonesia and Germany had already acted against the project before that. Governments, not critics, decided the biometric record was the greater risk.

A 2024 paper on personhood credentials, written partly by researchers at OpenAI, made the alternative explicit: prove you are human without revealing who you are, and not necessarily through biometrics at all. The choice, stripped down, is between two costs — certifying the body and risking the record, or labelling the work and risking that the label is ignored.

8. A performance that stopped being enough

Before any of this, there was already a theory for why presence used to be enough. Erving Goffman, in The Presentation of Self in Everyday Life, described ordinary life as a performance staged for others who read its cues — a glance, a pause, a tone — and judged a person real by how well those cues held together.

Goffman's whole argument depended on the performance being hard to fake. Once it stops being hard — once a hesitation, a stammer, a face's small involuntary movements can be generated as easily as recorded — the performance stops functioning as proof. What is left to check is not the cues anymore. It is the credential behind them.

9. Back in the room

Return to the meeting where this started. The badge beside the name has not changed how anyone in that call talks, looks, or argues. It has changed what the room is relying on to believe them. For most of history, a person's presence was the proof. Now presence is not enough; the room asks to see the badge instead, and the badge belongs to someone else.

How we read this

This essay draws only on the facts supplied in the brainstorm it was built from — Zoom's Verified Human badge and World ID, the decomp.dev closure, Reddit's phased API shutdown, the Wesleyan Media Project and New York Times figures on AI campaign ads, Thailand's order against World, the 2024 personhood-credentials paper, the CAPTCHA's history, and Goffman's 1959 work — plus the three signals supplied as evidence, cited to their original outlets. No scene, witness, quotation, score or figure beyond those sources was invented. Company claims (Zoom's on-device matching, World's anonymity-by-design) are reported as claims, not verified facts. The one figure that would most change this read is Tools for Humanity's actual enterprise revenue from World ID adoption, which remains undisclosed and would show whether certifying personhood is becoming a real business or still a cost centre dressed as one.

Grounded in Folka’s corpus: Proof of Human · The Deepfake Dividend · The Trust Recession