Definition

What is Proof of personhood?

Proof of personhood is Proof of personhood is a way to show that an online account belongs to one real, unique human, not a bot or a duplicate, without necessarily revealing who that person is.

Proof of personhood is a family of methods for answering one question online: is there a real, unique human behind this account? It does not have to say who the person is. It only has to show that they exist and that they are not using many accounts.

Why it exists

The internet was built so that anyone could create an account. That worked until software could fill in forms, write convincing text and pass for a person at huge scale. A fake review, a ballot-stuffing campaign or a free-trial abuse costs an attacker almost nothing when bots do the work. Services need some way to tell people from automation.

How it is done

Several approaches exist, and each trades privacy against certainty.

CAPTCHAs ask you to solve a puzzle. They are easy to use but increasingly solvable by AI.
Government ID checks compare a document to your face. They are strong but reveal your identity and are tied to one country's paperwork.
Biometric proofs scan something unique to your body, such as an iris, and issue a credential. They aim for uniqueness, but they raise privacy and consent questions.
Cryptographic credentials let you prove a fact, like "I am a verified human", without handing over the underlying data. They often use zero-knowledge proofs.

No method is perfect. Each can be gamed, can exclude some people, or can collect more data than users want.

The agent twist

AI agents change the question. If an agent acts for you, a service may want proof that a real person stands behind it, and that the person authorized the action. Proposals here let you delegate a proof to an agent, so a site can accept the agent while knowing a human is accountable. This sits alongside problems like prompt injection and browser agents being mistaken for attackers.

Proof of personhood is not the same as authentication, which proves you are the owner of one specific account. It also differs from content labeling such as AI watermarking, which marks machine-made output rather than verifying people.

What to watch

When a service asks you to prove you are human, check what it keeps. Ask whether it stores your biometric or ID data or only a yes or no, who can see it, and what happens if the company is breached. A strong claim of privacy is worth less than a clear description of the data.

Questions people ask