In a world where everything can be faked, how do you establish trust? Modem partnered with Swiss design studio Retinaa to develop a vision for proving personhood in the age of AI. Quartz is a ring-based ledger that pairs modern cryptography with the oldest technologies of trust humans have: meeting in person.
Beneath the quartz-made identifier sits enterprise-grade security: cryptographic keys stored on-ring, paired uniquely to its user via finger-vein biometrics.
As societies grow larger and more complex, their methods of establishing trust — in information, transactions and each other — become increasingly abstract. Although we may not see shaking hands, bowing or kisses on the cheek as being in any way related to facial recognition or cryptographic keys, these ancient gestures can be understood as the first forms in a genealogy of methods — what we might call “technologies of trust” — which enabled familiarity to spread through growing populations.
Historically, offering your hand to a stranger meant disabling the arm you used in combat. Bowing required you to lower your eyes and expose your neck, making yourself vulnerable in order to generate trust. When we stand opposite someone, or offer our cheek, we can feel reasonably confident about the intentions and the identity of the person we are dealing with. But as humans outgrew their tribal origins and began to form larger and more intricate communities, we needed to find new ways of scaling confidence and reassurance across ever-greater distances.
Technologies of trust beyond the body include things like wax seals, badges, uniforms and stamps, physical instantiations that represent a person, their family or a larger institution. Over time, as this technical apparatus grew, becoming increasingly computerised in the 20th century, new tools for forgery, deception, and counterfeiting were devised.
In summer 2025 a woman named Sharon Brightwell picked up her phone and heard her daughter April crying on the line. April had been texting while driving and had crashed her car into a young woman who was seven months pregnant. April had been arrested, she told her mother, and needed $15,000 to pay her bail.
“There is nobody that could convince me that it wasn’t her,” Brightwell later told reporters. “I know my daughter’s cry.” With remarkable haste, a bail bondsman showed up at the family home to collect the cash. Once he had left, Sharon received a second call from April. She told her that the woman’s baby had died and she was now threatening to sue. At this point Brightwell called a friend who told her she had likely been targeted by a new kind of scam.
Today deepfakes and AI-generated spoofs are cheap, quick and accessible. A system developed by Microsoft researchers can produce an accurate voice clone with just 3 seconds of sample audio. A new range of human weaknesses are being exploited.
Sharon Brightwell’s attackers used a voice clone of her daughter trained using material scraped from Facebook. Viewed from outside, and with the benefit of hindsight, the situation never quite added up (the “bail bondsman” was later revealed to be an innocent Uber courier). But when the situation is urgent enough, and especially when it involves a close family member, any one of us would be for forgiven for failing to remain sceptical.
Each stone contains natural crystalline inclusions forged over millions of years inside the Swiss Alps, producing a "chaos cryptography" so complex and specific that no computer algorithm or manufacturing process can duplicate it.
In the digital age, technologies of trust are based on code and mathematics, unintelligible to the average user. A digital signature is theoretically more secure than a handwritten one — but it is also less meaningful. Its ease of completion can lead to mistrust.
FaceID and multi-factor authentication are the contemporary equivalents of hand-drawn signatures and wax seals that had to be broken to access the information inside. Advanced encryption and reputation systems (ratings, reviews, scores) were designed to replicate the social behaviour of smaller human groups only now within the exponentially larger forum of the public internet — without any of the vulnerability and intimacy that made them work.
By meeting in person and shaking hands, the Quartz ring generates a Shared Secret via NFC, creating an unbreakable, encrypted bond between two people.
The challenge of verifying that a digital account is controlled by a real, unique human being, rather than a bot, autonomous agent, or human owner with multiple accounts, is known as the “unique-human problem” or “proof of personhood.” In an essay on the subject, Ethereum co-founder Vitalik Buterin proposes that a human ID system could incorporate specialised biometric hardware akin to the notorious orbs developed by Sam Altman’s former startup World. He suggests that in addition these biometrics could be used to generate multiple anonymous accounts for users, warding off the possibility that all online activity becomes traceable — tantamount to a mandatory digital ID.
“There is no ideal form of proof of personhood. Instead, we have at least three different paradigms of approaches that all have their own unique strengths and weaknesses … What we should ideally do is treat these three techniques as complementary, and combine them all.”
These techniques include (i) social graphs in which verified accounts vouch for and verify newcomers, (ii) general-purpose biometrics, like fingerprinting or FaceID, and (iii) specialised, more secure hardware, like the orbs. This arrangement has the potential to transform the digital sphere — everything from trustworthy online reviews to direct democracy and secure business interactions.
In August 2023 attendees at the DEF CON hacker conference in Las Vegas burst into applause as a live feed of Gal Zror — an IT vulnerability specialist — was transformed into the shape and voice of conference host Jeff Moss. Just a few weeks earlier, Zror had been tasked with creating a realistic deepfake of his boss at the identity security firm CyberArk. “There I was,” company CEO Udi Mokady told Bloomberg News, “crouched over in a hoodie with my office in the background.”
The deepfake wasn’t used to deceive employees at CyberArk. “If we were the size of an IBM or a Walmart or almost any Fortune 500 company,” Mokady added, “maybe Employee No. 30,005 could be tricked.” It was less than a year before an employee with the huge engineering firm Arup transferred 23 million euros into five fraudulent bank accounts after a conversation with a deepfake they had believed was the multinational’s CFO.
For now, creating a suitably realistic deepfake requires advanced chips, time and insider knowledge. But as with all kinds of software, it won’t be long before accessibility and efficiency improves. The threat is well understood by cybersecurity professionals and the clock is ticking. High profile executives inadvertently provide bad actors with ample training data: press photos, earnings calls, panel appearances and radio interviews. According to the business advisor firm Gartner, 62 percent of organisations experienced a deepfake attack in the 12 months leading up to September 2025.
The same tech striking fear into the hearts of corporate security teams can just as easily find us at home. A survey by the UK’s Starling Bank revealed 28% of UK adults had been targeted by an AI voice cloning scam. The figure is similar in the Netherlands. The slogan “every family needs a code word” is being pushed by concerned NGOs — a simple approach that relies on intimate contact. Even in a professional world mediated by digital communications we still meet in person to secure major business deals. We clink our drinks as we bond over food, not because we suspect we may be poisoned, but to observe one another’s humanity through universal rituals.
It’s likely a plurality of devices and approaches will be deployed to try and solve proof of personhood, each of which stands to learn something from the nuance and sophistication of gestures refined over thousands of years. Narrowing the gap between what is technically possible, and what feels trustworthy, will be an essential consideration as our technologies of trust continue to evolve.
The Verification: Transforming physical meetings into digital trust, Quartz uses zero-knowledge proofs to confirm "who you see" online matches "who you speak" to.
A): Onboarding
The user is prompted to wear the ring, which creates a cryptographic template based on the veins inside their finger.
Biometric pairing begins straight out of the box. To onboard and pair the ring, the user must place the ring on any finger of either hand.
During this first meeting, a cryptographic template is made by scanning blood flow through the veins inside the user’s finger. This template is a unique hash which cannot be reverse-engineered. It is stored on the device in a Secure Element inspired by similar isolated enclosures used for biometric security features like Face ID and Optic ID ⁵ in modern consumer tech. Crucially, no raw biometric data ever leaves the ring.
If the user has any concerns about the authenticity of the ring — a tampered-with box, or late delivery, for example — it can be taken to an External Verifier, authorised by the manufacturer. Each ring contains a quartz stone forged by nature over millions of years inside the Swiss Alps. The gemstone’s natural microstructures are turned into verifiable identifiers by recording its unique signature into an immutable blockchain-based digital certificate.
B): Pairing
Ring holders meet and shake hands while wearing their rings to produce a Shared Secret: a digital key for their relationship.
Two protocol users meet in person to forge a connection. By shaking hands while wearing their rings, they bring the devices into Near Field Communication (NFC) range, much like scanning a door pass or paying a cashier using your phone. Before shaking, both users press a small button on the side of the ring. This triggers the creation of a cryptographic key, which becomes part of a unique Shared Secret ⁷ for this specific 1:1 relationship.
From this point forward, if either ring holder suspects the connection may have been compromised — due to a missing ring, for example — the protocol mandates a new in-person meeting to establish a new Shared Secret.
C): Verification
When two users connect, a verification gateway confirms the presence of two connected rings worn by their rightful owners.
When the user wants to verify their contact is who they say they are, a gateway application will open, functioning as a seal which blocks access to calls, emails, and messaging platforms until it is unlocked. The gateway uses zero-knowledge proof (ZKP) technology ⁷ to confirm the user’s ring contains the Shared Secret specific to the requested connection without knowing who is being connected; the result is verification without a password being sent over the internet.
While the ZKP is performed, the ring performs a dual-factor “liveness check” ¹² during this verification window, which confirms the ring is being worn by the right owner (and that their finger is showing a pulse.) If the ring is removed or detected on a non-living object, verification automatically fails.As a final security feature, the gateway employs a “suspicion lock.” If the software detects hints that a spoof may be present, the call is blocked, though this can be overridden with a physical slow triple button tap from the user under suspicion to confirm a human presence.
Quartz
Vision, Research, Creative Direction: Modem + Retinaa
Writing, Research: Philip Maughan
Jewellery Design: Alicia Rosselet
Mineral Prospecting + Gem Cutting: Charles Centhuit
Lifestyle Photography: Matthew Tammaro
Product Photography: Robin Bucher
Photoshoot Production: CURRENT Production
Post Production: Purple Martin
Visual Identity & illustrations: Retinaa