Proof of personhood protocols jockey to establish networks of ‘verified humans’

Fake people are everywhere. Deepfake identities are applying for jobs, waging political campaigns, seeking romantic partners, slinging opinions online and, as of this week, testifying in court from beyond the grave.

Witness a recent incident on social forum Reddit, which saw a team of researchers release what TechCrunch describes as “a swarm of AI-powered, human-impersonating bots” on the “Change My View” subreddit. The team said they wanted to see how persuasive AI could be. The bots posted over 1,700 comments, adopting identities that displayed racist traits.

The sudden influx of so-called AI slop put Reddit into panic mode. This week, a post from the company CEO proclaimed that “Reddit works because it’s human” and announced that the site is exploring working with third party services that “verify humanity” –  the burgeoning market of providers offering proof of personhood (PoP) protocols, i.e. the digital assurance that one is not a bot.

In the dialectics of what we think of as “tech,” a problem created with code requires a response created with code. As fake people proliferate online and our longstanding definitions of “real” feel the tremors caused by generative AI, the quest to assure others that there is flesh and blood behind a username is about to hit the mainstream. And it could change not just how we interact online, but how we come to see ourselves in relation to the machines we have developed.

World U.S. launch comes with ambitions and baggage

Recently, in an Arizona courtroom, a deepfake recreation of a deceased man read his victim statement to the person who shot him dead in a road rage incident. About a week later in California, right around when the story his the news cycle, World – the proof of personhood company co-founded by Sam Altman, CEO of OpenAI and a key contributor to ChatGPT – held a livestream to announce its arrival in the U.S., bearing rectangular Orbs and fresh ambitions for a “global network of verified humans.”

World wants to scan the iris biometrics of anyone who’s interested, to register them in the World Network as proven humans. Its messianic branding matches its evangelical message: we can protect you from a world overrun by AI (which we may or may not have facilitated in the first place).

Yet the project also comes to the U.S. dragging behind it, like tin cans on strings, a chorus of global regulatory objections to its iris scanning practices. Thus far (as World is keen to point out) it has only been fully banned in one country: Hong Kong. However, the remaining objections are hardly trifling. Countries as far and wide as Germany, Kenya, Brazil and Indonesia have thrown shade of various depths at the company; the latter is the latest to suspend its biometrics operations, following public complaints. The likelihood of its U.S. operations going unchallenged in court seems low.

Regardless, it is the giant in the roomful of PoP vendors. It has Altman’s heft and reach, a tech and PR partner in the somewhat mysterious Tools for Humanity, a developing platform that already includes cryptocurrency, a wallet, apps and messaging – and a head start on most of its competitors in the space. Google “proof of personhood,” and the AI overview tells you proof of personhood is “also known as Proof of Human” – World’s inane, semantically void sobriquet for PoP.

This week saw PoP startup Humanity Protocol enter its Testnet Beta, paving the way for it to roll out its own biometric element: palm scanning. It has plans that match the size and scale of World’s. Other firms warming up on the sidelines include Civic Technologies, HumanCode, IDNTTY, the Holonym Foundation and Privado. The latter recently launched Billions, a non-biometric, decentralized PoP initiative which it calls “the first global human and AI network.”

“Billions revolutionizes verification by moving away from invasive biometric scans,” says a release. “Instead, users verify their identity just with their passport and phone.”

And yet a quote from co-founder Evin McMullen places Billions and its mission firmly within the emerging general model for PoP: “The fundamental mission of Billions.Network is to enable users to prove their humanity, access a plethora of benefits, and provide legal, safe proof-of-uniqueness anytime, anywhere, and for everyone.”

Everyone verified everywhere all the time

Common ideas underpinning PoP tech are that it is fully anonymous in allowing people to prove their humanness, that it gives users control of their data, and that absolutely everyone should do it. The first is rooted in blockchain technology and cryptographic zero-knowledge proofs (ZKPs), the second in data sovereignty.

The third – that PoP is necessary for everyone, as soon as possible – is taken as a given, because PoP protocols’ capacity to scale collapses without it. What should concern regulators, digital trust advocates and others in the biometrics space is that it might become true – not because the world is destined to see AI integrated into every aspect of our lives, but because those who develop it have convinced us it is so.

Or, it may be true already. Reddit has lived and died on anonymity, having historically required almost no personal information for users to create an account. That it is concerned enough about AI bots to modify its requirements is a troubling sign that the deepfake barbarians may already have breached the gate.

Article Topics

AI slop  |  blockchain  |  deepfake detection  |  deepfakes  |  digital trust  |  identity verification  |  proof of personhood  |  Reddit  |  social media  |  zero knowledge