Online protocol: If you’re not a bot, submit your proof-of-personhood to establish it

Subscribe to enjoy similar stories. Here we are in the third decade of the 21st century, and the internet has an identity crisis. For years, we worried about anonymity, privacy and whether our passwords were strong enough.

Now the more awkward question is whether the replies we get online are from a person at all. Generative AI systems can write posts, argue politics, flirt, invest, complain and even scam us with unnerving fluency. Digital avatars are eerily human.

The result is an online world where trust is evaporating because the basic assumption of a human on the other side of an interaction no longer holds. Proof-of-personhood has emerged as an attempt to fix this problem by giving the internet a way to distinguish humans from machines without turning everyone into a permanently monitored data point. Traditional digital identity systems ask you to prove that you control a credential, such as a password, phone number or government-issued document.

Proof-of-personhood asks a more basic question. Are you a unique living human being and not a script, model or copy of a scalable synthetic identity? This distinction matters because the economics of online abuse is built on scale. One human can type only so fast, but a single bot operator can create a million convincing accounts before breakfast.

As AI systems become cheaper and better, the cost of pretending to be human trends towards zero while the damage caused by that pretence rises sharply. We are moving from a world dominated by human-created content to one suffused with machine-generated speech, images and behaviour. As most online activity can be automated, systems that assume human scarcity start failing in unexpected ways.