Swift’s Blockchain Ledger Goes Live: What 17 Banks Are Actually Testing
For fifty years, Swift’s job was simple: tell banks where the money should go, then get out of the way. On July 9, 2026, the job changed. A blockchain ledger now sits inside Swift’s infrastructure, and seventeen banks across six continents are running real payments through it. The interesting part isn’t the blockchain. It’s what Swift decided not to touch.
Nine Months From Concept to Controlled Go-Live
Nine months is fast for a legacy financial network to move from an idea to a live pilot. Swift floated the shared-ledger concept at Sibos on September 29, 2025, promising more than 30 institutions would help design it, with Consensys behind the initial prototype. The build reached MVP stage on March 30, 2026, running an EVM-compatible architecture on Hyperledger Besu. On July 9, Swift called it ready for initial use and named seventeen banks preparing to pilot live transactions.
ANZ, BNP Paribas, BNY, Citi, DBS, First Abu Dhabi Bank, FirstRand, HSBC, Itaú Unibanco, Lloyds, Mashreq, MUFG, OCBC, Standard Chartered, UBS, UOB and Wells Fargo make up the roster, spread, Swift says, across six continents. Large global players sit alongside custody specialists and regional banks already running established tokenisation programs. The mix reads like a genuine cross-section, not a handpicked showcase. Swift is careful with the label too: a controlled go-live, not a commercial rollout, with more functionality promised once the phase proves out.
An Orchestration Layer, Not a New Currency
So what did Swift turn on? Not a currency. Swift issues no money on the new ledger and has built no cryptocurrency. Banks in the pilot issue tokenised deposits, digital stand-ins for a customer’s existing deposit claim, on ledgers they run individually. Swift’s layer records, sequences and validates the payment commitments moving between banks, then lets smart contracts enforce the rules.
Here’s the distinction worth sitting with: a tokenised deposit never leaves the regulated banking system. A bank issues it, carries it on the balance sheet, and runs it through the same deposit, compliance and risk framework covering every other dollar it holds. A stablecoin usually sits outside the banking system entirely, backed by reserve assets and circulating on public blockchains through wallets and exchanges no bank controls. Swift’s design keeps the money inside a perimeter regulators already know how to police.
Thierry Chilosi, Swift’s chief business officer, calls it an extension of “the trust and stability of established finance” into digital money, with an eye toward programmable-money and agentic-commerce use cases down the line. HSBC, which plans to plug its Tokenised Deposit Service into the ledger, pitches the appeal in blunter terms: real-time payments across time zones, no artificial cut-offs.
Ethereum Under the Hood, With Bank-Grade Guardrails
Under the hood, the MVP runs on Ethereum Virtual Machine-compatible technology built on Hyperledger Besu, an enterprise Ethereum client built for permissioned networks. Consensys backed the original 2025 prototype, though Swift’s July release stops short of confirming whether the firm still holds the primary technology contract.
The setup does not put Swift on Ethereum’s public mainnet. No bank is buying Ether to make a payment. The permissioned design borrows smart-contract programmability while keeping participation, governance and data visibility locked to vetted institutions. Real questions remain open: who runs the validating nodes, how data gets partitioned between banks, what happens when a smart contract needs a rewrite. The answers will matter more than the architecture once the pilot moves past its first transactions.
Around-the-Clock Access Is Not Instant Settlement
Real-time. Instant. Always-on. Swift, HSBC and the rest of the pilot banks reach for the same words, and none of them has published a service-level agreement, a benchmark, or a single corridor-specific timing figure to back it up. The vocabulary describes ambition, not a measurement.
Three clocks run at once here, and mixing them up produces a rosier picture than the facts support. The ledger validates and records a payment commitment fast. A receiving bank can release customer funds outside normal banking hours, overnight, on a weekend. Final interbank settlement, meanwhile, still runs through existing systems in the current design, and nothing in Swift’s release guarantees the settlement step happens in seconds.
The gap between funds showing up and the debt clearing for real raises a question every treasury team should ask before it gets excited: who eats the intraday or weekend credit exposure if a payment moves before settlement finishes. Nobody has answered it publicly yet.
Swift, Ripple, Agorá and Fnality Are Racing Toward the Same Goal on Different Roads
Swift isn’t the only outfit chasing always-on cross-border payments, and the differences between the four leading attempts say more than any of the marketing decks.
| Model | Primary money form | Approach | Current status |
| Swift shared ledger | Bank-issued tokenised deposits | Orchestration layer connected to existing rails | Controlled initial go-live |
| Ripple Payments | Stablecoins, XRP, fiat | Managed digital-asset payment network | Commercial service |
| Project Agorá | Tokenised deposits and central-bank reserves | Shared programmable wholesale platform | Prototype moving to real-value testing |
| Fnality | Central-bank-backed digital cash | Currency-specific DLT payment systems | Sterling system live; broader network developing |
Ripple sells stablecoins, XRP and fiat through a managed payment network built outside traditional banking rails, with XRP Ledger transactions Ripple says settle in roughly three to five seconds. Project Agorá, run by the Bank for International Settlements alongside the Federal Reserve Bank of New York and more than 40 institutions, demonstrated atomic multi-currency settlement combining tokenised central-bank reserves with commercial-bank deposits in a May 2026 prototype, still short of production. Fnality backs its digital cash directly with funds parked at central banks, tying settlement finality to central-bank money instead of orchestration between commercial banks.
None of the three substitutes for what Swift built. Ripple bets on speed and reach outside the existing banking system. Agorá bets on redesigning the settlement layer itself. Fnality bets on tying digital cash to central-bank money. Swift’s bet is the most conservative of the four: connect the tokenised deposits banks already plan to issue, and ask none of them to give up the compliance machinery or Swift connectivity they already run on. In banking, conservative bets tend to outlast bold ones.
The Payoff for Treasury Teams Depends on What Happens After the Demo
The upgrade doesn’t matter to a corporate treasurer chasing a faster Monday-morning wire. It matters to one running Friday-afternoon payroll for an office opening Monday in a different time zone. Overnight and weekend fund movement, faster intercompany transfers, and a clearer read on cash positions all become real once a receiving bank can release funds outside conventional banking hours.
Banks won’t benefit evenly. Global transaction banks already issuing tokenised deposits, HSBC and Citi among them, sit closest to the center of the pilot. Correspondent banks living on fees from slow, multi-step payment chains face the opposite exposure, and so do fintechs whose entire pitch has been speed. None of it plays out until the pilot produces real transaction volume instead of a handful of showcase payments.
What Swift Hasn’t Published Will Decide Whether the Ledger Scales
Swift has not said when the first live transaction will happen, which corridors or currencies it covers, what volume it expects, or how governance works in practice. The price tag is a blank too, so nobody outside the seventeen banks can say whether the ledger costs less to run than the correspondent-banking chain it’s supposed to replace parts of. Regulatory treatment adds a fog on top: no single authority signs off on a cross-border bank ledger, and central banks, prudential supervisors and data-protection regulators across multiple jurisdictions will each claim a piece of it.
Swift has tried blockchain before without any of it sticking: a 2017 nostro-reconciliation proof of concept, CBDC interoperability work in 2022 and 2023, neither one scaling past the lab. What separates the current attempt is the roster, seventeen live participants, and the nine-month sprint from concept to controlled use. Momentum isn’t proof.
Swift hasn’t moved global banking onto a public blockchain, and it hasn’t replaced the settlement systems underneath cross-border payments. What it’s built is a bridge: regulated deposits move with fewer time-zone constraints, while banks keep the legal and operational scaffolding they already trust. The next year will show whether seventeen banks turn a pilot into real volume, or whether blockchain’s most credible banking experiment stalls exactly where every one before it stalled: at proof of concept.
