Worldcoin is launching by giving as many people as possible a share of a new currency.

Imagine a world where everyone around the globe, regardless of who they are, could participate in the growing digital economy and benefit from decentralized, collective ownership. What would it take to get there?

The first step is getting the same currency into the hands of as many people as possible. The more people there are that collectively hold and use the same currency, the more useful that currency becomes for each participant. While it’s not hard to see that widespread adoption is desirable, every currency faces a coordination problem on the way: it is difficult to convince new users to adopt a currency that is not widely held, but it is hard to reach widespread adoption unless enough users participate.

Largely for this reason, crypto as a technology has only reached about 3% of the world’s population to date^[1]. While there have been amazing technological breakthroughs, there has yet to be a successful, concerted effort to achieve widespread, global adoption.

Worldcoin aims to overcome this coordination problem by aligning the incentives of all participants and allocating the majority of the currency to as many unique new users as possible as an incentive to join the network.

Distinguishing between new users in a way that protects their privacy and can scale to the whole world is a hard problem. Our approach relies on a custom biometric device - we call it the Orb - that verifies the uniqueness of a person through iris recognition, while ensuring their privacy through zero-knowledge cryptography.

There were many technical challenges along the way. In the following sections, we describe our design decisions for different aspects of the problem including biometrics, hardware, crypto and economics. But to start, we want to offer a glimpse into what the global launch of Worldcoin might look like by providing the results from our field tests, which have been underway for months now!

It is core to the mission of Worldcoin that the network is not built and run by a single entity, but instead by entrepreneurial individuals all around the world who apply to receive an Orb to run independent operations in their communities. We call these individuals Orb Operators and give them the autonomy to build their operations from the ground up, doing everything from running their own marketing campaigns to collaborating with local partners. 

In order to test this model and technology, we have onboarded twenty-five Operators that run more than thirty devices in twelve countries across four continents (Africa, South America, Europe and Asia). In the coming months, the scale of these tests will increase as hundreds more devices and Orb Operators come online. After launch and an initial Orb production ramp-up phase, around four thousand devices will be distributed per month.

So far, the best performing Operators have been able to onboard around one thousand users per Orb per week, with our best performing Operator even signing up 2,198 users in one week with a single Orb. They have achieved these results despite using prototype devices (which are slower than the production model), facing frequent COVID disruptions, and not having the benefit of a public Worldcoin presence. In the four weeks before this post was published, the average weekly sign-up rate across all Operators was 764 per Orb and 1,069 per Operator. 

To illustrate the success and potential of these Operators, we will dive deeper into the operations from the following five countries: Chile, Kenya, Indonesia, Sudan and France.

Using the numbers we have seen in the field, we decided to conduct a thought experiment to make these results more tangible and help visualize what could happen with increased Orb production. The figure below demonstrates how cumulative sign-ups over time could increase significantly as more Orbs come online. Although we anticipate that there will be all kinds of saturation effects and other complications, we believe that there is a path for this model to onboard billions of users.

We are thrilled by these results—they show that our strategy for getting crypto into the hands of as many people as possible just might work. Of course, these are still early days and achieving widespread adoption is really just the beginning, but these results are certainly promising. 

In the following sections, we want to present the technical challenges that we needed to solve in order to get to this point.

The launch of Worldcoin requires a robust and scalable way to determine whether someone is unique (in a sense of new) and actually human (not a bot). So-called Proof-of-Personhood protocols^{[2, 3]} try to solve this problem by allowing people to prove they are “unique humans”^[4]. This “unique-human proof” would unlock many meaningful applications^[5], but none of the existing approaches have yet reached a global scale.

The all-too-familiar CAPTCHA^[7] was one of the first approaches towards  a Proof-of-Personhood for the internet. It is a test designed to be difficult for bots, but easy for humans.  While it does well in verifying humanness, CAPTCHA fails to verify uniqueness: as a human, you can pass as many tests as you want. This could be addressed by requiring everybody to solve a CAPTCHA challenge at the same time^[8], but that would be very difficult to coordinate on a large scale.

Many other approaches to Proof-of-Personhood^{[9, 10, 11]} have been deployed. Some of these use social verification to ensure accounts are owned by unique humans. This often requires that existing users meet (physically or virtually) new users and deposit some form of financial stake that can be taken away in case of fraud. Social verification systems may lead to a robust Proof-of-Personhood, but they also face difficult scaling challenges.

Biometric approaches (e.g., Aadhaar in India) are promising as they are both accessible and enable accurately verifying uniqueness. The key challenges for biometric approaches are related to privacy, fraud detection, and scalability. We believe that recent advances in technology offer a way to solve these issues.

Worldcoin leverages biometrics, cryptography, and incentives to create an elegant, scalable, and privacy-preserving solution to the “unique-human problem”. The following sections describe how the system works and details how we plan to deploy it globally.

Verification & Identification

At their core, biometric systems usually consist of:

• A device that acquires biometric information (e.g., fingerprints) and converts them to a compressed representation (an “embedding”); and
• A database that stores these embeddings for later retrieval and comparison.

Embeddings are designed so that one can calculate a distance (in the sense of similarity) between two embeddings. If this distance is above a certain threshold, the two embeddings are assumed to belong to different people. 

One typically distinguishes between two modes of operation for biometric systems: verification and identification. 

Verification is the most familiar use case of biometrics. For example, access to a building may require a keycard and a fingerprint scan. Here, fingerprint verification helps ensure that a stolen keycard cannot be used to enter the building. Behind the scenes, the keycard is linked to an embedding in the database, which is compared to the one of the scanned finger.

In the case of identification, there is no keycard available to help: the biometric system has to determine a user’s identity from the biometric scan alone. A common use case for biometric identification is in forensics where, for example, a fingerprint or DNA sample is compared with every entry in a database to identify a suspect.

Biometric Proof-of-Personhood

A biometric system can be used as the foundation for Proof-of-Personhood by ensuring each person can only register once. Similar to biometric identification, the system needs to compare each user’s embedding to every other entry in the database. If none match, the user’s registration is successful and their embedding is added to the database. If there is a match, the user has likely registered before and is therefore rejected. 

However, there is a significant challenge with this approach: the identification problem becomes increasingly difficult with a growing database. Issues with search complexity and performance come into play, but more importantly, the system starts making more errors while trying to match users. These errors are largely governed by the “false match rate” (FMR). The FMR measures the likelihood that two biometric scans are incorrectly considered the same. At the scale of billions of users, the FMR must be extremely small to avoid frequently rejecting legitimate new users.

The key to being able to scale such a system to onboard billions of people is capturing an information-rich biometric source at a high level of fidelity. The more information the source contains, the lower the probability of false matches.

Iris Recognition Ensures Uniqueness, Fraud Resistance, And Practicality

We evaluated many different technologies and believe that iris recognition is the best-suited for our biometric approach to Proof-of-Personhood.Compared to other biometric methods, like face recognition, fingerprints and DNA sequencing, iris recognition stands out by simultaneously satisfying the following three criteria:

1. Uniqueness: The iris theoretically contains enough information to distinguish nearly all humans uniquely.
2. Fraud-resistance: The iris is difficult to physically modify and the biometric capture can be made hard to spoof.
3. Practicality: The required imaging technology already exists, and for the user, presenting their iris is a fast and simple process.

On a global scale, a biometric uniqueness check must confidently distinguish each person from every other person. The information density of the iris is high enough to allow this, and no two irises are alike (even for identical twins). It also turns out that it is possible to capture most of this information with high-resolution imaging of the iris. 

The same is not true for today’s face recognition technology: at the scale of billions of people, there is simply not enough information in, for example, a selfie to reliably tell people apart. Even advanced algorithms such as neural networks can’t help. 

The information of the iris pattern is distilled into an embedding, in our case the so-called IrisHash, which is derived from a one-way function with the iris image as the input. This code is stored in a database and is the only data used for checking uniqueness. Under this approach, the original image of the iris does not need to be stored anywhere.

We describe this in more detail in our post on privacy.

Using the iris also enables a strong degree of fraud-resistance. In the context of our biometric uniqueness check, this means we need to reliably detect if someone attempts to spoof the biometric capture (e.g., using animals or special contact lenses). Since cheating a uniqueness check only requires looking different than any existing user, the fraud resistance property of the iris is critical. The iris is inherently hard to modify and enables better ways to verify liveness compared to other common biometrics like faces and fingerprints.

Finally, any scalable uniqueness check must be fast and simple. From the user’s perspective, presenting an iris is as easy as posing for a photo. In fact, the practicality of iris recognition has already been proven at scale. One notable example is the border control system for the United Arab Emirates, which leverages iris recognition^[13].. Some other biometric approaches, like DNA sequencing, may not provide the same degree of simplicity for the user, and can feel invasive.

Using iris recognition for a scalable Proof-of-Personhood requires an iris scanner that’s secure, high-resolution, and easy-to-use. We searched for existing devices that meet these requirements at a global scale, but found none. For this reason, Worldcoin developed the Orb.

High Resolution Imaging Is Key For Biometric Systems At Scale

The primary function of the Orb is to capture detailed images of a person’s eyes, and convert them to an IrisHash. From the outside, the UX is simple: users simply look into the camera (which is behind the black surfaces) and press a button. On the inside of the Orb, a custom optical system automatically finds the user's eyes and brings them into focus. The Orb then captures high-resolution images of both irises at multiple infrared wavelengths, then uses them to locally compute the user’s IrisHashes.

To work at scale, the Orb must reliably capture high-quality iris images across a variety of lighting conditions and distances from the user. This adaptability is achieved using a combination of custom optics hardware and machine learning. To quickly find the user’s eyes, the Orb uses a movable mirror that is guided by a neural network. Once the eye is in-frame, another neural network controls a custom liquid lens to focus on the detailed patterns in the user’s iris. All of this happens while the Orb’s infrared LEDs are adaptively illuminating the eye, ensuring consistent images across lighting environments.

Together, these systems enable the Orb to capture irises with significantly higher quality: up to two orders of magnitude higher resolution compared to the industry standard^{[14, 15]}. At this level of detail, the unique patterns in the iris can be used to reliably distinguish among billions of people.

The Orb’s Security Measures Resist Fraudulent Activity

Security is core to the Orb's design. After all, Worldcoin requires trusting that the IrisHashes are legitimately produced by an Orb. The Orb’s security system ensures the authenticity of each IrisHash by robustly detecting a wide range of fraudulent behaviors. Broadly, we can classify such behavior as either spoofing or tampering.

Spoofing attacks involve presenting the Orb with modified, fake, or non-human irises. For example, an attacker might show the Orb a photo of an iris or an animal iris, hoping to generate a unique IrisHash. To defend against attacks like this, we’ve equipped the Orb with a suite of multi-spectral sensors and custom fraud-detection algorithms. This advanced anti-spoofing system complements the iris imaging system, and operates locally on each device. 

The Orb is also resilient to various forms of tampering, including attempts to modify its software, extract its cryptographic secrets, or disable its anti-spoofing system. This resilience is critical, since any of these intrusions might allow a hacker to generate fraudulent IrisHashes. The Orb’s embedded systems reliably detect advanced attacks of this kind, and prevent corresponding fraud.

To further increase the difficulty of an attack, Orbs will be remotely monitored and compared to other Orbs. Such monitoring is based on non-biometric metadata from the Orb, including battery level, temperature, and network strength. Anomalies will be flagged and lead to Orbs being deactivated. This anomaly detection happens in the cloud and therefore comes with higher security guarantees than device-level spoof and tamper detection.

While the Orbs are important to solve the unique-human problem, it must also be ensured that those Orbs are used in the most effective way. Therefore, each Orb will be handled by the Orb Operator whose task it is to explain Worldcoin to new users and to operate the Orb to help them sign up. Operators earn Worldcoin for each sign-up, which incentivizes them to maximize their sign-up rate (i.e., the number of sign-ups per week). Combined with the proof-of-personhood, this makes Worldcoin the first cryptocurrency to directly encode, into its trustless protocol, the economic incentives for users to join the network: giving every new user a share of its currency.

Still, one challenge is to find Operators that can actually achieve a high sign-up rate. This can be done at scale through a bidding process that allocates Orbs to the most efficient Operators. Concretely, Operators bid for the right to operate one or multiple Orbs for a defined period of time. Operators who sign up more users than others will achieve a higher monthly revenue, so they can be expected to submit higher bids, making them more likely to win the bidding process. Additional incentive structures are in place for promoting not only raw number of sign-ups, but also high-engagement of users and disincentivizing fraud. We will share more details about those mechanisms in the future.

The described incentive mechanism is core to Worldcoin. Of the fixed supply of 10 billion tokens, the vast majority will be given to users and Operators that are helping to launch Worldcoin through the protocol. To incentivize early adoption of the currency, the number of tokens that users and Operators receive per sign-up will decrease over time. Less than twenty percent of the total supply of tokens will be used to fund Orb production and initial protocol development. Three percent will be allocated to continued ecosystem development and maintenance (like grant issuances, for example).

With the Orb, Worldcoin aims to onboard the first billion users into the world of cryptocurrency. We believe Worldcoin can reach this unprecedented scale by leveraging recent innovations in blockchain scalability and embracing compatibility with the existing crypto ecosystem. This section details our approach, including the mechanisms for ensuring user privacy once our token is live.

Scaling & Compatibility

Worldcoin’s scalability is provided by an optimistic rollup^{[16, 17]} on top of Ethereum. This "layer 2" scaling solution ensures transactions will remain cheap, fast, and reliable as the Worldcoin network grows. Importantly, this solution maintains the open and permissionless nature of Ethereum: anybody can create accounts, submit transactions, and participate in validation. The optimistic rollup is built using the open-source Hubble project^[18], extended with a high-performance sequencer implementation that will be open-sourced in the coming weeks.

Worldcoin aims for broad compatibility. The underlying token is based on Ethereum’s popular ERC-20 standard, enabling compatibility with many existing tools and services. In addition, our open-source SDKs (see Wallet & SDK) will make it very easy for developers to interact with the currency and the Layer 2. 

Users are free to withdraw their funds to Ethereum mainnet or migrate through bridges to any Layer 2 on Ethereum (e.g., Optimism, Arbitrum, and many others) or even separate and aspiring ecosystems, such as Solana or Polkadot.

Account Privacy

Worldcoin inherits Ethereum's privacy model: accounts are pseudonymous and transactions are public. Crucially, Worldcoin accounts are never associated with any biometric data from the Orb. Zero-knowledge proofs (ZKPs) are the enabling technology behind this privacy-preserving biometric approach. We use the open-source Semaphore^[19] zero-knowledge proof system to transfer the uniqueness of IrisHashes to the uniqueness of user accounts, without ever linking them.

Phase 1: Sign-up and uniqueness check

• The user generates a Semaphore keypair on her phone.
• The user presents the hashed public key (via QR code) to the Orb.
• The Orb scans the user’s irises and locally computes the user’s IrisHash.
• The Orb sends a signed message containing the hashed public key and the IrisHash to the sign-up sequencer node.
• The sequencer node verifies the Orb’s signature, then checks if the IrisHash does not match any already in the database. If the uniqueness check passes, the IrisHash and the public key are saved.

Phase 2: Claiming Worldcoin

• The user’s app locally generates a wallet address.
• The app uses Semaphore to prove that it owns the private counterpart to one public key registered in Phase 1. Because it’s a zero-knowledge proof, it does not reveal which public key.
• The proof is sent again to the sequencer, which verifies it and initiates the deposit of tokens to the provided wallet address.
• A so-called nullifier is sent along with the proof and ensures the user cannot claim the reward twice.

Those two steps - decoupled by the zero-knowledge proof - enable a biometric-based Proof-of-Personhood without linking user accounts to biometric data. A more visual explanation of account privacy can be found in our dedicated privacy blog post.

Worldcoin is designed to be the first widely-adopted cryptocurrency. Making this a reality means going further than having an inclusive launch: Worldcoin must be useful and understandable to people with no prior experience using digital money. In addition, users must be confident in the privacy and security of their accounts. Because of this, we are developing a mobile wallet application.

The wallet allows users to learn about Worldcoin and other cryptocurrencies, find Orb Operators, claim their share, view their balance, send global peer-to-peer payments, and use their Worldcoin throughout the Ethereum ecosystem. No central authority mediates transactions; users always have full control and ownership of their Worldcoin (i.e., the wallet is “non-custodial”), which they can back up if they choose.

In addition to the wallet, we are developing a collection of open-source Software Development Kit (SDKs) to help developers build products around Worldcoin—things like alternative clients, tools for niche use-cases, digital tipping, in-store payments, and professional interfaces. These SDKs, coupled with pre-existing decentralized financial protocols, will allow anyone to serve the needs of local, regional, and global markets.

Most excitingly, these SDKs will eventually include zero-knowledge Proof-of-Personhood for third-party applications, maintaining user privacy as described in the Account Privacy section. While we built the Orb to help make Worldcoin a reality, we know this primitive can unlock much more. That’s why we want to give everyone the ability to reuse their already issued proofs to access other projects.

We believe the wallet and SDKs will help make Worldcoin — and the larger crypto ecosystem — more accessible and useful. We’re excited to share more details about these tools over the coming months.

There are many ways in which Worldcoin can have substantial positive effects on our society. Still, we strongly believe that how we accomplish this project is just as important as what we achieve. Our design approach is based as much on our own deeply-held values as on technical reasoning. We are choosing to prioritize user privacy, data security, openness, and broad accessibility, even when it means tackling some of today's toughest engineering and operational challenges.

For example, building on Ethereum at this scale is more challenging than operating a custom, centralized blockchain, at least in the short-term. The same applies to the privacy approach using zero-knowledge proofs, which is on the edge of current research. But we think this is the right way to launch Worldcoin, and the effort will pay off in the long run for all of us.

If successful, Worldcoin will, within the first year, onboard hundreds of millions of new people who didn’t have access to this technology. This will hopefully result in many developers integrating Worldcoin into new and existing products. We believe the internet has entered a new era: Web3, where finally economic value and ownership are natively integrated into the web, and proof of personhood is critical. It is hard to imagine what new products this coming era will bring. Certainly, it will involve wider adoption of novel technologies and models of collaboration, like DAOs, NFTs, and decentralized exchanges.

There’s still a lot of work to do to enable and support a rapidly sprawling ecosystem of developers, entrepreneurs, and community builders. We are in the early stages of designing Worldcoin’s long-term governance strategy, grant programs, and partnerships roadmap. We are committed to following credible neutrality^[20] as a key guiding principle in setting up these mechanisms. 

There are still many challenges left, and we will continue our research on many of them, such as blockchain scalability, digital identity, and pseudonymous economies.

On a final note, while we are deeply committed to giving Worldcoin to everyone on Earth, legal uncertainty in some places means we will not deploy Orbs in these countries right away. That said, we will continue to explore ways to make this possible, including pushing for clearer rules, so that we can give Worldcoin to everyone, regardless of who they are or where they are from.

Join us and let’s build together!