Online, an account looks like a person. But an account is not proof of a person. Software and coordinated operators can create, control, and amplify identities at a scale ordinary humans cannot match.
Hash Humanity starts with a different question: who is actually human?
Hash Humanity is built around proof of personhood. Before an account becomes another voice in the network, HumanKey is designed to establish that a real, live, unique person stands behind it.
On conventional platforms, another account may require little more than another email address, phone number, automated registration workflow, or compromised credential. When identities are disposable, abuse becomes inexpensive to repeat.
A person has limited time and attention. Software does not operate under the same constraint. Automated systems can create posts, replies, follows, reactions, messages, and other account activity continuously and at large scale.
Likes, replies, reposts, follows, and repeated talking points can create the appearance of popularity or consensus even when a much smaller number of operators are coordinating the activity behind them.
When an abusive identity can immediately be replaced, moderation becomes an endless cycle of account removal while the underlying operator simply returns under another identity.
When genuine people cannot reliably be distinguished from automated identities, coordinated accounts, impersonators, and synthetic activity, trust in the conversation itself begins to erode.
Automation, account farms, generative AI, stolen credentials, and coordinated operators can combine to perform activity that once required large numbers of people.
Automated identities can follow accounts, react to posts, repost material, generate comments, and repeatedly amplify the same message, creating activity that appears organic.
Coordinated networks can repeat narratives, overwhelm replies, create apparent agreement, and make a fringe position appear substantially more popular than it really is.
Fake accounts can copy names, photographs, branding, writing styles, or organizational identities and use those identities for fraud, misinformation, phishing, or social engineering.
Automated systems can send messages, post content, scrape information, test responses, and reach large audiences much faster than a single human operator could manually.
Generative systems can produce large amounts of unique text, imagery, audio, and other synthetic material, reducing repetitive patterns that once made some automated campaigns easier to recognize.
If account creation remains cheap, deleting a malicious account addresses the symptom rather than the operator. Another identity can simply take its place.
Microsoft reported blocking an average of roughly 1.6 million bot-driven or fake-account signup attempts every hour during the period covered by its 2025 Digital Defense Report. The point is larger than one company: automated identity creation operates at industrial scale.
Microsoft Digital Defense Report 2025 ↗HumanKey is designed around one human, one account. If every additional verified account requires another unique human instead of another script, disposable email address, or automated registration, large-scale manipulation becomes substantially more difficult and expensive to organize.
HumanKey is Hash Humanity's proof-of-personhood system. Its job is not to learn who you are. Its job is to establish that a real, live person is requesting access and help determine whether that person has already enrolled.
HumanKey begins after the user agrees to the verification process. Camera access and biometric processing are not intended to occur silently in the background.
A liveness challenge helps determine whether the camera is seeing a live participant rather than a static photograph, basic replay, or other presentation attack.
HumanKey derives a mathematical facial representation suitable for comparison. Its purpose is duplicate-person detection, not creating a public identity profile.
The representation can be compared against enrolled representations to determine whether the same person appears to have already created an account. This supports the one-human, one-account model.
After successful verification, HumanKey can use cryptographic proofs and derived identifiers so the network can recognize that personhood requirements were satisfied without requiring a public legal name or unnecessary identity information.
After enrollment, routine authentication can use a passkey instead of repeatedly scanning the user's face. Passkeys use public-key cryptography and avoid relying on a traditional reusable password.
The goal is not continuous facial verification. Biometric verification is intended for enrollment and controlled recovery scenarios where HumanKey must re-establish the returning person.
Software can create unlimited processes. It cannot create unlimited unique human beings. Requiring unique personhood places a physical constraint in front of digital account creation.
If one person cannot simply create another verified identity, losing an account carries more consequence. That can make bans, rate limits, and reputation systems more meaningful.
A verified account does not prove that a person is truthful, wise, or acting in good faith. It does provide stronger evidence that a distinct human stands behind the account.
Coordinated humans can still organize, but requiring another verified person for another account makes it much harder for one operator or automated system to cheaply multiply itself into an artificial crowd.
A person can choose not to reveal a civil identity while still proving that a unique human stands behind an account. Personhood and public identity are different things.
When identities are harder to throw away and replace, behavior can carry history. What a person builds, contributes, and says can have continuing meaning.
Proof of personhood should not require turning a social network into an identity database. HumanKey is designed around data minimization: process what is necessary to establish personhood and uniqueness, then rely on secure credentials and privacy-preserving proofs for ordinary access.
Users can participate without publishing a government name or other civil identity information.
Passkeys are intended to handle routine authentication after successful HumanKey enrollment.
The biometric system exists to establish personhood and uniqueness, not to track a person's activity around the network.
HumanKey proves personhood, not character. It cannot determine whether a statement is true. It cannot prevent humans from coordinating, lying, harassing, manipulating, or exercising bad judgment.
What it can do is reduce one of the internet's most powerful force multipliers: the ability for one operator to cheaply masquerade as an unlimited number of independent people. Hash Humanity is not trying to create a network where everyone agrees. It is trying to create one where disagreement is actually happening between people.
Hash Humanity is an experiment in what social media can become when the network begins with a scarce resource software cannot manufacture: a real human being.
