Inside Colossus’ Ambitious Plan to Dethrone Visa and Mastercard With KYC-Free Crypto Cards
Joseph Delong’s desk no longer resembles the neat setup of a typical software engineer. Instead, it looks like the backroom of a small electronics workshop: scattered card readers, dismantled terminals, cables coiled around prototype hardware, and stacks of dense manufacturer catalogs.
For Delong, a veteran Ethereum developer and former CTO of SushiSwap, this chaos is not clutter-it’s research. All of it is feeding into his latest and most audacious project: Colossus, a stablecoin-based credit card network built on an Ethereum layer-2 that aims to sidestep both banks and the traditional payment giants.
He jokes about his ever-growing “box of goodies”-a collection of point-of-sale test terminals, magnetic-stripe and NFC readers, chip card prototypes, and obscure reference devices that most engineers never see. But getting that hardware, he says, has been one of the hardest parts of building a challenger to Visa and Mastercard.
“Accessing this stuff feels like trying to break into an old guild,” Delong explains from his home office in San Antonio, Texas. “The knowledge is arcane, the supply chains are closed, and most of the documentation assumes you’re already an established bank or payment processor. It’s not designed for outsiders.”
Colossus is very much an outsider. The company is tiny-just four people-but it’s targeting one of the most entrenched and regulated infrastructures in the world: card payments. The team plans to launch its Ethereum layer-2 network in March, positioning it as a sovereign settlement layer that replaces bank rails with crypto-native infrastructure.
At the core of Colossus is a simple but radical promise: “Not your keys, not your card.” Instead of linking a plastic card to a bank account, Colossus ties it directly to a crypto wallet that users control themselves. The system is designed to work without traditional Know Your Customer (KYC) checks, relying instead on on-chain logic and collateralization to manage risk and payments.
Where today’s card networks rely on a mesh of issuing banks, acquiring banks, payment processors, and card schemes, Colossus wants to collapse that stack into a software-driven protocol. Settlement, authorization, and risk management would move from proprietary bank databases to smart contracts running on an Ethereum-compatible layer-2.
The design is built around stablecoins rather than volatile cryptocurrencies. When a user swipes or taps a Colossus card, the goal is for the merchant to effectively receive stable-value funds, while the underlying settlement occurs in a crypto-native way. That’s how the project hopes to make crypto invisible at the checkout counter while still giving users full self-custody and bypassing traditional banking rails.
Delong’s workbench full of terminals and readers underscores a crucial point: hardware is the choke point. Legacy card networks have spent decades locking down standards for EMV chips, magnetic stripes, NFC payments, and terminal certifications. For a new entrant, simply getting development access to terminals and understanding how they communicate is a major hurdle.
“Most of this industry runs on standards that are public in theory but closed in practice,” Delong says. “If you’re not already inside the ecosystem, you don’t know who to talk to, what documents you need, or which vendor will even return your call. We had to reverse-engineer not just the tech, but the business relationships behind it.”
Colossus’ answer is to meet the existing system where it is. The goal is not to replace physical cards or payment terminals outright, but to speak their language. The network’s cards should behave to merchants’ terminals like any other card: you tap or insert, the terminal runs its usual checks, and the transaction is approved or declined. Under the hood, however, the approval is anchored to an Ethereum layer-2, and the settlement is stablecoin-based rather than bank-based.
By doing that, Colossus is trying to avoid the classic crypto problem at point-of-sale: merchants don’t want volatility, unfamiliar processes, or new hardware. They want interchange fees they understand and chargeback rules they can predict. For consumers, a Colossus card is supposed to feel indistinguishable from a traditional debit or credit card-except that the balance is held in a wallet, not in a bank.
The KYC angle is where the project becomes especially controversial. Traditional card issuance is tightly linked to identity verification regulations. Colossus is pushing a model where cards can be issued and used without standard KYC checks, because the “account” is effectively a crypto address backed by on-chain collateral rather than a bank account owned by a named customer.
To manage risk, the protocol can enforce rules like collateral ratios, spending limits, and transaction constraints algorithmically, rather than tying them to a person’s legal identity. In theory, the card becomes a programmable credit line linked to crypto assets, with the network acting more like an automated clearing and risk engine than a bank.
This, however, immediately raises regulatory questions. Even if Colossus doesn’t see itself as a bank, regulators may view any system that issues payment cards and touches fiat on- or off-ramps as part of the traditional financial perimeter. Delong appears to be betting that by anchoring the system in crypto-native rails and self-custody, Colossus can carve out a distinct regulatory niche, or at least move faster than incumbents bound by legacy infrastructure.
The Ethereum layer-2 component is key. A base-layer blockchain like Ethereum is secure but often too slow and expensive for high-frequency, low-margin payment traffic. Card networks process tens of thousands of transactions per second at peak; a naive on-chain equivalent would be unusable for everyday retail. A specialized layer-2 can batch transactions, optimize data formats, and handle the throughput needed for card volume, while periodically settling back to Ethereum for security.
Colossus is designing its L2 to function less like a generalized smart contract playground and more like a purpose-built payments fabric. That includes predictable fees, near-instant finality at the user level, and APIs that resemble what traditional payment processors already use. In other words, developers integrating Colossus shouldn’t need to think in terms of gas, nonce management, or transaction ordering; they should be able to send payment instructions and get confirmations.
Stablecoins also play a dual role in the architecture. On one side, they are the unit of account that merchants see-something close to a digital dollar, euro, or other fiat-denominated token. On the other, they are the collateral and settlement medium inside the network. This allows Colossus to sit at the intersection of DeFi liquidity and real-world commerce: the same stablecoins circulating in lending pools or DEXes can back a user’s spending limit on a card.
Yet the decentralization trade-offs are real. To interface with the highly regulated card ecosystem, Colossus will likely need some centralized or semi-centralized components: issuing entities, custodial fallback mechanisms, and dispute-handling processes. Balancing those realities with the ethos of “not your keys, not your card” will be an ongoing tension. Users may love the idea of pure self-custody, but merchants and payment partners demand reliability, fraud controls, and clear responsibility when things go wrong.
Security is another frontier. A typical cardholder today can lose their card or have it skimmed, and banks will often absorb fraud or reverse charges. In a self-custodial crypto-card model, who ultimately eats the loss? Colossus will need to design robust safeguards: multi-factor authorization for high-risk transactions, dynamic spending profiles, and possibly on-chain insurance pools that can compensate in case of theft or protocol bugs.
For users, the appeal is obvious: a payment card that connects directly to crypto holdings without surrendering control to a bank or centralized exchange. They could spend stablecoins at any merchant that accepts standard cards, without preloading funds into a custodial app or relying on intermediaries that can freeze accounts. For people in regions with capital controls or weak banking infrastructure, that kind of direct, protocol-level access to a global payments network would be transformative.
For merchants, the selling point is more nuanced. Colossus could potentially reduce fees by compressing the traditional stack of issuers, acquirers, and processors into a slimmer protocol layer. Smart-contract-based settlement can also shorten settlement times, improving cash flow. And because transactions are crypto-native, it may open up new types of loyalty, cross-border commerce, and programmable invoices that aren’t possible on legacy systems.
However, adoption will hinge on more than just technology. Visa and Mastercard enjoy deep trust, legal clarity, and decades of compliance frameworks. Colossus must convince partners that its model can handle disputes, refunds, fraud investigations, and regulatory audits with the same rigor. That means a mix of transparent on-chain data and off-chain governance-something many crypto projects still struggle to design.
The project also exists against a rapidly evolving backdrop: major card networks themselves are experimenting with stablecoin settlement, crypto-linked cards, and tokenized deposits. Banks are exploring their own blockchain-based rails. Colossus is not just competing with the incumbent card brands of yesterday but with their future, more crypto-aware versions.
Still, Delong believes that starting from first principles in crypto offers an advantage. Where incumbents are layering tokenization on top of decades-old infrastructure, Colossus can design every component-authorization flows, settlement batches, risk models-around a blockchain-native mindset. That could allow for faster iteration, simpler integrations, and more flexible products for both consumers and businesses.
Inside that “box of goodies” on Delong’s desk lies the tangible evidence of just how much friction stands between crypto and everyday payments. Each terminal he dissects is a reminder that payments is a hardware problem as much as a software one, entangled with proprietary standards and long-standing commercial arrangements. Colossus’ bet is that by embracing those constraints instead of ignoring them, it can smuggle a new, decentralized settlement layer into the heart of the global card network.
Whether a four-person team can meaningfully challenge the financial giants that process trillions of dollars each year remains an open question. But if Colossus’ Ethereum layer-2 does launch as planned and gains traction, it will offer a live test of one of crypto’s oldest ambitions: replacing-or at least routing around-the legacy infrastructure of Visa and Mastercard with a system where users hold the keys, the protocol holds the rules, and the card is just the interface.