ECB Project Agora: What Tokenised Cross-Border Settlement Means for EU Payment Firms

ByOfficer

Last updated: May 2026

On 27 May 2026, seven central banks and more than 40 financial institutions published the findings of Project Agora, a two-year prototype exploring tokenised cross-border settlement of wholesale payments. The Bank of France represented the Eurosystem alongside the Bank of England, the Federal Reserve Bank of New York, the Bank of Japan, the Bank of Korea, the Bank of Mexico, and the Swiss National Bank. The project, convened by the Bank for International Settlements (BIS) and the Institute of International Finance (IIF), demonstrated that tokenised central bank reserves and tokenised commercial bank deposits can settle cross-border wholesale payments atomically, across currencies and jurisdictions, on a shared programmable platform.

That sentence sounds theoretical until you consider what it replaces. Today, a EUR-to-JPY wholesale payment typically passes through three or four correspondent banks, each running its own sanctions screening, each holding liquidity in separate pools, each operating on different schedules. Project Agora showed a working alternative: a single coordinated settlement where all legs complete or none do. For EU payment firms, banks, and PSPs watching how their settlement infrastructure might change over the next five to ten years, this is the first large-scale prototype to demonstrate that the approach works within existing legal frameworks.

Related reading: EPC Q1 2026 Payments Reports Roundup: What Matters for EU Payment Firms

What Project Agora Actually Built

Project Agora is not a product and not a payment system. It is a prototype built on a unified ledger concept first proposed by the BIS in its 2023 Annual Economic Report. The concept involves bringing tokenised central bank reserves and commercial bank deposits onto a shared distributed ledger technology (DLT) platform, enabling multi-currency settlement through smart contracts.

The prototype has a two-layer architecture. A unifying ledger records tokenised commercial bank deposits and is accessible to all direct participants. Below that, independent jurisdictional ledgers record tokenised central bank reserves, one per currency. Each central bank retains full autonomy over its own jurisdictional ledger, including reserve issuance, validation rules, and oversight arrangements. The design preserves jurisdictional sovereignty while enabling coordinated cross-border settlement.

Direct participation is limited to central banks and commercial banks. Non-bank financial institutions can participate as initiation service providers (maintaining a node on the unifying ledger to initiate payment-versus-payment transactions on behalf of participating banks) or as indirect participants (holding deposit accounts at participating commercial banks, with balances recorded as tokenised deposits). No retail participants, no individuals. This is wholesale infrastructure.

One common misunderstanding I see in payment operations discussions is treating tokenisation projects as if they create new financial instruments. The Project Agora legal analysis is explicit: tokenised reserves and tokenised deposits do not alter the legal characterisation of central bank reserves or commercial bank deposits. The issuance of tokenised reserves and deposits does not create new assets. The underlying account relationships, rights, and obligations remain the same. The token is a record-keeping mechanism on a shared ledger, not a new instrument class.

Atomic Settlement: What It Changes for Payment Operations

The central operational innovation is atomic settlement. In the prototype, all balance updates across the affected unifying and jurisdictional ledgers execute as a single coordinated outcome: commit (settle) or cancel (abort). There is no partial movement of funds across currencies and jurisdictions.

For teams managing wholesale cross-border payment flows today, the operational pain is familiar. Sequential processing means liquidity gets committed before downstream checks are complete. A payment fails at the third correspondent bank, and now you are unwinding transactions, reconciling across three time zones, and managing the credit exposure created by partial settlement. The BIS report identifies these as structural pain points: failed payments, reconciliation burdens, siloed liquidity, and misaligned operating hours.

The prototype addresses each of these differently from current infrastructure. Compliance checks, including sanctions screening, AML/CFT controls, and confirmation of payee, happen before liquidity is committed. The platform brings forward the alignment of payment-related information so that those processes occur prior to the commitment of liquidity and settlement. Settlement then occurs in seconds once funds are locked. The platform is designed to operate around the clock, which eliminates the friction caused by misaligned cut-off times between jurisdictions.

The real problem starts when payment operations teams ask: does atomic settlement actually remove settlement risk? In the prototype, yes. The report states that atomic settlement is achievable in tokenised central bank reserves and tokenised commercial bank deposits across all seven participating jurisdictions. But the report also distinguishes between atomicity in the workflow sense (coordinator-governed commit-or-cancel across legs) and legal settlement finality. The legal analysis found that settlement finality is achievable across all seven jurisdictions, but further work is needed to define the technical, operational, and contractual requirements best aligned with each jurisdiction’s legal framework.

The Legal Framework Question That Matters Most

The most significant finding for EU payment firms is the legal and regulatory analysis. The project examined how tokenised reserves and deposits interact with existing law across seven jurisdictions, including the EU (represented through France and the Eurosystem). The conclusion: the prototype does not present any direct conflicts with existing laws and regulations. Implementation of a platform with this design is achievable within the current legal and regulatory framework.

That is a strong statement, and it needs context. The analysis covers the legal nature of tokenised reserves and deposits, settlement finality, contractual frameworks, financial crime prevention compliance, data privacy and confidentiality, and data localisation. It found no unmanageable challenges. But it also identified areas where further elaboration is needed, including record-keeping, operational resilience, outsourcing oversight, deposit insurance applicability, customer confidentiality rules, and the contractual documentation governing platform participation.

For the EU specifically, the GDPR does not impose data localisation but regulates cross-border transfers of personal data. The report notes that the distributed storage of transaction records across nodes in different jurisdictions may raise data governance challenges. French policy discussions around critical payment data highlight potential scrutiny of where data are stored or accessed. These are not blockers, but they are the kind of operational questions that EU compliance teams will need to work through if a production platform materialises.

Where teams commonly get this wrong is in assuming that a positive legal assessment for a prototype means regulatory clearance for a production system. It does not. The report explicitly states it does not make definitive legal determinations regarding how the prototype would necessarily operate in practice, nor does it make regulatory recommendations. The contractual documentation, rulebook provisions, and supervisory confirmations needed for a live system are all future work.

AML/CFT and Sanctions Screening on a Tokenised Platform

Financial crime compliance is where the operational model differs most from current correspondent banking. In the prototype, each institution implements AML/CFT controls and sanctions screening independently and confidentially. The platform ensures that all required validations are completed prior to settlement, but the checks themselves stay within each institution’s own systems.

The prototype incorporates global payment data standards, including legal entity identifiers (LEIs) and ISO 20022 CBPR+ messaging, and includes a confirmation-of-payee process. The confirmation of payee is privacy-preserving: the creditor institution verifies beneficiary details within its own systems, and only the minimum endorsed outcome is shared for workflow progression.

The report also notes that the modular design could enable enhanced financial crime-related information-sharing in future, as regulatory and data-sharing frameworks evolve. The platform includes functionality that would support this, but the report is careful not to overclaim: it acknowledges that current tipping-off rules in certain jurisdictions may require additional caution when designing information-sharing mechanisms.

I find the approach sensible for an institutional prototype, but payment compliance teams should note one gap. The report’s AML/CFT analysis examines the existing framework. It does not account for the EU’s transfer of AML/CFT functions from the EBA to the new Anti-Money Laundering Authority (AMLA), which took effect on 31 December 2025. AMLA is now the responsible body for AML/CFT standards in the EU, including the draft AML/CFT regulatory technical standards it consulted on from 9 February 2026. Any production system touching EU jurisdictions would need to track AMLA’s evolving rulebook, not the EBA’s legacy framework.

Where This Sits Alongside the Digital Euro and ECB Payment Infrastructure

Project Agora is one of several ECB-adjacent payment infrastructure initiatives. It is worth mapping them clearly, because the field is getting crowded and the overlap potential creates confusion.

The digital euro is a retail central bank digital currency under development by the ECB, with a PSP pilot open call and scheme rulebook consultations running in parallel. It targets person-to-person and person-to-merchant payments. Project Agora is wholesale only. They serve different layers of the payment stack.

Pontes is the ECB’s own project for linking DLT-based market infrastructures with TARGET Services. The ECB roadmap places Pontes pilot launch at Q3 2026. Pontes focuses on securities settlement and collateral management using DLT, while Agora focuses on wholesale cross-border payments.

T2 extended operating hours are a separate but related initiative, with the ECB exploring how to extend TARGET Services availability to address the same time-zone friction that Agora tackles through 24/7 platform design.

The SEPA instant payments regulation addresses intra-EU retail instant payments, while PSD3 overhauls the payment services framework. Both operate in the retail space.

The practical takeaway: these initiatives address different segments, but they share common infrastructure questions around settlement finality, data governance, and AML/CFT integration. Firms mapping their payment infrastructure strategy should track them as a portfolio, not in isolation.

What EU Payment Firms Should Not Expect

This is a prototype, and the BIS is direct about it. Project Agora is not about building a finished product. It has delivered a prototype to evaluate the potential of this new approach. The project intends to advance testing, including conducting real-value transactions involving certain currencies and participants. The Bank of Canada has also joined.

There is no production timeline. There is no commitment from any participating central bank to deploy a production system based on this design. The report does not recommend that firms begin building interfaces to a tokenised settlement platform.

What the report does say is that future work is expected to involve an enhanced role for the private sector, supported by continued and active engagement from participating central banks. The project will advance to real-value testing. If you are a large EU bank already participating through the IIF consortium, you know what that means for your team’s allocation. If you are a smaller PSP or payment institution, the operational signal is softer: this is directional, not imminent.

It is easy to read too much into a prototype. A prototype demonstrating feasibility is not a regulatory mandate. No EU regulation currently requires tokenised settlement capability. PSD2 and the forthcoming PSD3 do not reference tokenised reserves. The Settlement Finality Directive (Directive 98/26/EC) governs designated payment and securities settlement systems. Whether a future production platform based on Agora’s design would need designation under the SFD is an open question the report flags but does not resolve.

Liquidity and Treasury Implications

For bank treasury and liquidity operations, the prototype offers a specific promise: reduced liquidity fragmentation. In the current correspondent banking model, banks pre-fund nostro accounts across multiple jurisdictions, with each balance sitting idle until a payment arrives. The report identifies siloed liquidity as a structural pain point that complicates cash and treasury management.

Atomic settlement on a shared platform changes the liquidity dynamic. Because all legs of a cross-border payment settle simultaneously, there is no need to pre-position funds across multiple intermediaries. The platform’s path discovery mechanism identifies available routes before liquidity is committed. Combined with 24/7 availability, this could allow banks to manage cross-border liquidity with fewer pre-funded balances and tighter intraday controls.

The limitation the report acknowledges is that the prototype’s liquidity model has not been tested under stress conditions or at production volume. Intraday liquidity behaviour on a tokenised platform with atomic settlement is different from the current model. Banks subject to BCBS 248 intraday liquidity monitoring would need to evaluate how their monitoring frameworks apply to a platform where settlement occurs in seconds rather than in batch cycles. This is not an ALMM reporting change. EBA ALMM templates C 66.00 to C 71.00 capture maturity ladder, funding concentration, and counterbalancing capacity at the reporting reference date. They do not capture real-time settlement flows. But the operational inputs feeding those reports would look different on a tokenised platform.

Governance, Rulebook, and Contractual Requirements

The report identifies a set of governance and contractual requirements that would need to be established before any production deployment. These include: permitted assets and transaction types, settlement finality definitions and legal effect, accounting and risk management approaches, exception handling procedures, liability and compliance standards, data governance requirements, reporting and record-keeping standards, and legal framework provisions including governing law and dispute resolution.

For EU-supervised institutions, the governance overlay matters. The platform’s direct participants are central banks and commercial banks. Commercial banks that issue tokenised deposits retain all current regulatory and supervisory obligations. The platform does not create new regulatory exemptions or lighter-touch oversight. If anything, the multilateral governance structure of a shared ledger adds complexity around operational resilience, outsourcing oversight, and regulatory access to data.

The report notes that designation of components of the platform as a payment system may be required in at least some participating jurisdictions. For the EU, this could trigger Oversight Framework requirements under the ECB’s role as overseer of systemically important payment systems. It could also raise questions under the Markets in Financial Instruments Regulation (MiFIR) if the platform processes transactions that qualify as reportable under those regimes.

The regulatory surface of a production system is easy to underestimate. The prototype operates within existing legal frameworks. A production system would trigger a distinct set of authorisation, oversight, and reporting questions that the prototype has not needed to resolve.

Frequently Asked Questions

What is Project Agora?

Project Agora is a public-private collaboration convened by the BIS and the IIF, involving seven central banks (including the Bank of France representing the Eurosystem) and more than 40 financial institutions. It built a prototype demonstrating how tokenised central bank reserves and commercial bank deposits can settle wholesale cross-border payments atomically on a shared DLT platform. The report was published on 27 May 2026.

Does Project Agora affect current EU payment regulations?

No. Project Agora is a prototype. It does not create new regulatory requirements. PSD2, the forthcoming PSD3, SEPA regulations, and the Settlement Finality Directive remain unchanged. The project’s legal analysis found no direct conflicts with existing laws, but a production system would need separate legal, contractual, and supervisory frameworks.

Who can participate in a Project Agora-type platform?

Direct participation is limited to central banks and commercial banks. Non-bank financial institutions can participate as initiation service providers (initiating payment-versus-payment transactions on behalf of banks) or as indirect participants (holding deposit accounts at participating banks). There is no retail participation. Individuals cannot participate.

What does atomic settlement mean for settlement risk?

Atomic settlement means all balance updates across currencies and jurisdictions complete or none do. This eliminates the credit and settlement risk that arises from partial completion in sequential correspondent banking chains. The prototype demonstrated this across seven jurisdictions. Legal settlement finality was assessed as achievable, but further work is needed on the contractual and operational requirements in each jurisdiction.

How does AML/CFT compliance work on the platform?

Each institution performs its own AML/CFT checks and sanctions screening independently and confidentially. The platform ensures all required validations are completed before settlement proceeds, but does not centralise compliance. The design incorporates LEI identifiers and ISO 20022 CBPR+ messaging standards. Enhanced information-sharing capabilities could be added as regulatory frameworks evolve.

Is there a production timeline?

No. The project will advance to real-value testing involving certain currencies and participants, and the Bank of Canada has joined. There is no commitment from any participating central bank to deploy a production system. Future work is expected to involve an enhanced private-sector role, but no dates have been announced.

What should EU banks and PSPs do now?

For most firms, this is a monitoring brief, not an action item. Large banks already in the IIF consortium should track the real-value testing phase. All EU payment firms should map Project Agora alongside the digital euro, Pontes, T2 extended hours, and PSD3 as part of their medium-term infrastructure planning. No system changes or compliance actions are required based on the prototype findings.

Related Articles

Key Takeaways

  • Project Agora demonstrated that tokenised central bank reserves and commercial bank deposits can settle wholesale cross-border payments atomically across seven jurisdictions. The report was published on 27 May 2026 by the BIS and IIF.
  • The prototype’s legal analysis found no direct conflicts with existing laws and regulations across the participating jurisdictions, including the EU through the Eurosystem. Implementation within current legal frameworks is assessed as achievable.
  • Tokenised reserves and deposits do not create new financial instruments. They are record-keeping mechanisms on a shared ledger. The legal nature of the underlying account relationships does not change.
  • Atomic settlement eliminates the credit and settlement risk from partial completion in sequential correspondent banking chains. Compliance checks, including AML/CFT and sanctions screening, occur before liquidity is committed.
  • There is no production timeline and no regulatory mandate for tokenised settlement. This is a prototype with plans for real-value testing. No EU regulation currently requires tokenised settlement capability.
  • EU payment firms should map Project Agora alongside the digital euro, Pontes, T2 extended hours, and PSD3 as part of medium-term infrastructure planning.
  • The governance, contractual, and supervisory frameworks needed for a production system have not been developed. Designation under the Settlement Finality Directive and ECB oversight frameworks are open questions.

Sources and References

Disclaimer: The information on RegReportingDesk.com is for educational and informational purposes only. It does not constitute legal, regulatory, tax, or compliance advice. Always consult your compliance officer, legal counsel, or the relevant supervisory authority for guidance specific to your institution.

Similar Posts