Not all biometrics are created equal. Not at a scale of 8 billion people, anyway. It’s important to understand why.
First, let’s start with some important definitions and context:
In the context of biometrics, the higher the entropy, the more unique and unpredictable the biometric characteristic. This entropy can be described in “bits”, or units of measurement used to quantify the amount of entropy present in a biometric sample.
With that groundwork established, let’s look at why biometrics are key to proving personhood.
The human iris is incredibly complex. It has hundreds of unique points like crypts, furrows and pigment spots. This complexity gives it very high entropy.
Studies have shown that the iris code generated from a scan of a human eye can have over 200 bits of entropy—significantly more than the entropy contained in a fingerprint scan and enough to theoretically distinguish between 8 billion individual humans. Similarly, iris recognition is around four orders of magnitude (10,000x) more accurate than facial recognition due to a higher entropy in the iris structure compared to the face. Iris recognition algorithms based on iris codes can even be used to distinguish between identical twins because iris textures stem from a random process during gestation and are not related to a human’s DNA.
It’s worth noting that, while these are educated estimates based on the inherent features of the different biometric markers, the performance of biometric systems can vary depending on many factors. Nevertheless, high entropy is one of the key reasons why iris biometrics are essential for verifying the personhood of everyone on Earth. But it’s not the only reason.
Beyond entropy, the human iris is protected within an internal organ (the eye), making it less susceptible to changes, alterations or variations from external factors. The texture of a healthy iris remains almost unchanged throughout an adult human’s life.
The same can’t be said for fingerprints or faces. Fingerprints, due to their exposure, can be more easily altered (intentionally or not) or worn over time than irises. Similarly, the reliability of facial recognition can vary dramatically based on conditions like facial expression, age and so on. Facial recognition algorithms are also more prone to bias than irises, with well-documented divergent error rates across demographic groups.
At a scale of billions of individuals, this increased susceptibility to change and elevated demographic bias are further reasons why fingerprints and facial recognition are unsuitable as a biometric solution to global proof of personhood—especially for a system dedicated to the prevention of duplicate sign-ups.
In addition to their entropy, reliability and inclusiveness, tech advancements over the past decade have made it practical to scale iris biometrics globally. This includes the Orb.
After studying various forms of biometrics, the team that set out to build Worldcoin spent years developing the Orb, a portable biometric imaging device whose hardware engineering files are progressively being released to enable others to develop, build and operate similar devices that integrate with the World ID protocol. The Orb has been customized to meet the stringent requirements of the Worldcoin protocol. These requirements include:
You can learn more about the Orb’s robust privacy features here or get more information on its hardware like a custom telephoto lens here.
Stay up to date on all of the latest Worldcoin announcements, events and information by signing up for the blog newsletter at the bottom of this page. You can also follow along on Twitter, Discord and Telegram.
The above content speaks only as of the date indicated. Further, it is subject to risks, uncertainties and assumptions, and so may be incorrect and may change without notice. A full Disclaimer can be found in our Terms of Use.