Saudi Arabia’s recently disclosed departure from Project mBridge has reopened a wider question about the future of cross-border payments. The Saudi Central Bank has clarified that its participation ended in May 2025 after completing a planned proof of concept, and that it did not join subsequent phases. On its own, that does not amount to a geopolitical realignment or a verdict on mBridge. But it brings a more structural question back into view.
For much of the past decade, experiments in central bank digital currencies (CBDCs), tokenised deposits and distributed ledgers have focused on whether technology can make cross-border payments faster, cheaper and more efficient. Increasingly, however, the more important question is no longer whether monetary systems can be connected. It is how those connections should be organised, and where authority over access, liquidity, data and settlement should sit once they are.
Project mBridge, Project Agorá and Project Nexus offer three different answers. All respond to the same problem: cross-border payments remain dependent on multiple intermediaries, fragmented systems and separate domestic infrastructures. But they differ substantially in how much of the existing monetary architecture they seek to change.
Payment infrastructure is not simply technical plumbing. The design of the rails can determine which institutions remain necessary, how national authorities retain control, where liquidity sits and who sets the rules through which money moves internationally. The emerging contest is therefore not simply between technologies. It is between different models of interoperability and different distributions of monetary power embedded within them.
The issue: connecting monetary systems without giving up monetary control
Traditional cross-border payments often move through chains of correspondent banks, foreign-exchange providers, messaging networks and domestic settlement systems. Each additional stage can introduce cost, delay, reconciliation requirements and settlement risk. CBDC and tokenisation experiments have shown that some of these steps can be compressed. Different ledgers can communicate, multiple currencies can settle atomically, and payment-versus-payment arrangements can connect both sides of a forex transaction more directly.
The harder problem is now institutional. As these systems move beyond experimentation, countries must decide how much infrastructure they are willing to share. Greater integration can reduce duplication and intermediary layers, but it also requires more questions to be resolved jointly: who can participate, how data is governed, where liquidity sits, and how compliance, settlement finality and operational responsibility are allocated. The trade-off is therefore straightforward: the more infrastructure is shared, the more governance must also be shared. mBridge, Agorá and Nexus sit at different points on that spectrum.
Three ways of solving the same problem
1. mBridge: common infrastructure for sovereign settlement assets
Project mBridge represents the most integrated model. Initially developed by the BIS Innovation Hub with the central banks and monetary authorities of China, Hong Kong, Thailand and the UAE, mBridge created a common distributed-ledger platform through which participating central banks could issue wholesale CBDCs and commercial banks could conduct cross-border payments and foreign-exchange transactions. Its appeal lies in shortening the conventional settlement chain. Rather than moving through separate national systems and multiple intermediaries, sovereign settlement assets can operate within a common environment. But deeper integration creates a deeper governance requirement.
When the BIS Innovation Hub handed mBridge over to its partner central banks in October 2024, after the project reached minimum viable product status, the institutional question became more visible: if several central banks share the settlement layer, how much governance must they share with it? Once sovereign currencies sit on common infrastructure, questions around participation, data, liquidity, currency controls and operational responsibility can no longer be answered entirely within national boundaries.
That is what makes mBridge significant beyond CBDCs themselves. The more integrated the settlement rail becomes, the more consequential the governance of that rail becomes.
2. Agorá: modernising the infrastructure while preserving the monetary hierarchy
Project Agorá starts from a different premise. Rather than placing several wholesale CBDCs on a common settlement layer, Agorá seeks to make the existing two-tier monetary system more programmable. Its prototype combines tokenised commercial-bank deposits with tokenised central-bank reserves. Commercial banks continue to issue the money used by their customers, while central banks retain control over reserves and final settlement. What changes is the infrastructure connecting those forms of money.
The BIS’s May 2026 results showed that the prototype could support atomic multi-currency settlement while allowing jurisdictions to retain control over their own central-bank reserves. Subsequent real-value testing involved 28 financial institutions and central banks across 17 transaction scenarios. Agorá therefore changes less about who is permitted to issue money and where settlement authority sits, and more about how existing forms of central bank and commercial bank money are represented, connected and settled across jurisdictions. The underlying two-tier monetary structure is preserved. The innovation lies in making those existing claims more interoperable and programmable.
3. Nexus: connecting systems rather than changing the settlement asset
Project Nexus takes a different approach again. Unlike mBridge and Agorá, Nexus does not depend on introducing a new form of central-bank money or tokenising the existing monetary system. Instead, it seeks to connect domestic instant-payment systems through a common multilateral framework. Under the model, a domestic payment-system operator connects once to Nexus and, through that connection, gains access to the other participating systems.
In 2025, the central banks of India, Indonesia, Malaysia, the Philippines, Singapore and Thailand incorporated Nexus Global Payments to move the model towards live implementation. The distinction is therefore clear: mBridge changes the settlement infrastructure; Agorá changes how existing monetary claims interact; Nexus changes the connections between domestic payment systems. That makes Nexus particularly important to the wider CBDC debate. It suggests that at least some of the gains associated with cross-border CBDCs – faster, more direct and less fragmented payments – may be achievable without introducing a new settlement asset at all.
The real contest is over interoperability
Looking at the three projects together changes how the transformation in cross-border payments should be understood. The debate is often framed as a technological race: CBDCs against tokenised deposits, or blockchain-based infrastructure against conventional payment systems. But the more important distinction is institutional. Each model places authority somewhere different.
mBridge embeds more coordination at the level of common settlement infrastructure. Participating jurisdictions may gain greater efficiency, but they must also develop shared rules around how that infrastructure operates. Agorá preserves more of the existing two-tier monetary architecture. Central banks retain control over their own reserves, while commercial banks retain their established role in issuing customer money. Coordination therefore takes place across sovereign monetary systems rather than through a single multi-CBDC settlement layer. Nexus requires even less change to the underlying monetary architecture. Countries can retain their domestic currencies and payment systems while adopting a common framework through which those systems communicate.
Power is therefore not removed, but relocated. Under mBridge, it increasingly sits in the governance of the common platform. Under Agorá, it remains distributed across national monetary systems but becomes dependent on common technical and operational arrangements. Under Nexus, influence can sit in standards: how systems connect, which specifications they follow and what participation rules govern the network.
This is why the debate cannot yet be reduced to competing geopolitical blocs. A simple bipolar narrative, in which mBridge becomes one financial ecosystem and Western-led tokenisation initiatives another, goes further than the evidence currently supports. mBridge itself has involved jurisdictions with different geopolitical alignments. The BIS described its 2024 withdrawal as the project’s graduation from the Innovation Hub, while Saudi Arabia has characterised its own participation as a planned proof of concept rather than a geopolitical reversal.
What the evidence does suggest is more nuanced: cross-border financial infrastructure is becoming a new site of monetary competition, but that competition has not yet hardened into closed systems. Instead, jurisdictions appear to be testing different ways of balancing three objectives that do not always sit comfortably together: greater efficiency, deeper interoperability and continued national control. That also means countries may not ultimately choose one architecture exclusively.
Multi-CBDC platforms may prove useful for particular wholesale transactions. Tokenised reserves and deposits may become relevant for institutional markets. Connected instant-payment systems may be better suited to retail payments and remittances. The likely end-state may therefore be a layered system in which several forms of cross-border infrastructure coexist. If so, the next challenge will not simply be interoperability within each new system. It will be interoperability between them.
Why Nexus makes this particularly relevant for India
For India, this debate is not abstract. Its participation in Nexus places it directly within one of the emerging models for reorganising cross-border payments. That is significant because Nexus offers a path to deeper international connectivity without requiring India to place the rupee onto a shared multi-CBDC settlement platform or fundamentally redesign its domestic monetary architecture. Instead, it builds on something India has already developed at considerable scale: instant domestic payment infrastructure. This gives India a different strategic entry point into the cross-border payments debate.
The question is not simply whether the digital rupee should eventually become interoperable with other CBDCs. It is also whether the objectives of faster payments, lower friction and more direct links between jurisdictions can be pursued by connecting existing payment systems. Nexus suggests that they can. But it also means India will confront the same governance questions visible in mBridge and Agorá in a different form. A network connecting national instant-payment systems still requires common standards, participation criteria, operating rules and arrangements for cross-border transactions. Monetary sovereignty may remain largely domestic, but the governance of connectivity becomes multilateral.
For India, therefore, the strategic question may not be whether it chooses CBDCs or payment-system interoperability. It may ultimately use both. The more consequential choice is which functions India is willing to place on shared infrastructure, which it wants to retain domestically, and how much influence it can exercise over the standards connecting the two. Saudi Arabia’s departure does not determine mBridge’s future. Agorá’s progress does not establish a universal alternative. Nexus does not make wholesale CBDCs redundant. Taken together, however, the three projects suggest that the technical challenge of connecting monetary systems is gradually giving way to a governance challenge.
The next phase of cross-border payments will therefore be shaped not simply by which rail processes transactions fastest, but by which model of interoperability countries are willing to trust, and how much monetary authority they are willing to embed in the connections between them.