How CBDC Pilots Are Reshaping Payment Infrastructure

Central bank digital currencies are moving from research papers to production infrastructure faster than most payment operators have adjusted their thinking. e-CNY (China’s digital yuan) reached 16.7 trillion yuan (approximately $2.3 trillion) in cumulative transaction volume by November 2025, more than an 800% increase from 2023 (People’s Bank of China, November 2025). mBridge — the multi-CBDC cross-border settlement platform developed by the BIS Innovation Hub, HKMA, Bank of Thailand, UAE Central Bank, and People’s Bank of China — reached Minimum Viable Product stage in June 2024, with the BIS handing operational control to partner central banks in October 2024 (BIS, October 2024). The ECB has completed its digital euro preparation phase and is advancing toward a legislative decision.

The question for payment operators is not whether CBDCs matter — they clearly do at this scale — but which CBDC developments are operationally relevant now versus which are long-cycle policy discussions that won’t affect infrastructure for 5–10 years. The answer requires separating the retail CBDC story (government-to-citizen) from the wholesale CBDC story (bank-to-bank, cross-border), because the two have entirely different infrastructure implications and timelines.

Wholesale CBDCs: Where the Infrastructure Is Being Built

Wholesale CBDCs — digital central bank money used exclusively for interbank settlement — are the most operationally significant development for payment operators in the near term. They don’t replace consumer payment methods; they change how banks settle between themselves, which affects settlement finality, FX transaction costs, and correspondent banking relationships.

mBridge: Multi-Currency Wholesale Settlement

mBridge is the most advanced multi-jurisdiction CBDC project in production. The consortium includes the HKMA (Hong Kong), Bank of Thailand, UAE Central Bank, the Digital Currency Research Institute of China’s PBoC, and Saudi Central Bank (joined as full participant, June 2024), with 26+ observing members. The BIS Innovation Hub handed operational control to the partner central banks in October 2024, a signal that the project has transitioned from research to live operation (BIS, October 2024).

The technical architecture: mBridge runs on a distributed ledger platform (mBridge Ledger, built on an Ethereum-derived consensus mechanism) that allows participating central banks to issue CBDC tokens representing their respective currencies. Cross-border payments between participating jurisdictions settle by swapping CBDC tokens directly on the ledger, without a correspondent bank intermediary.

The elimination of correspondent banking in these corridors has direct cost implications. SWIFT-based cross-border transactions in Asian corridors typically cost $25–$50 per transaction in correspondent banking fees, plus 0.5–2.0% FX markup, plus 2–4 business days for settlement. mBridge settlement is atomic (simultaneous exchange of currencies), settling in seconds at a fraction of the cost.

The current scope is limited to participating jurisdictions and financial institutions that have completed the connectivity and compliance requirements. For operators with USD-to-AED, USD-to-HKD, USD-to-THB, or USD-to-CNH payment flows through participating banks, mBridge-enabled settlement corridors will become available as participating banks complete their mBridge integrations. The timeline for operator-accessible corridors is 2026–2027.

Project Dunbar and the Interoperability Architecture

Project Dunbar (BIS Innovation Hub, Reserve Bank of Australia, MAS, South African Reserve Bank, Bank Negara Malaysia) explored multi-CBDC settlement with a different design philosophy: rather than a single shared ledger, Dunbar tested connecting separate national CBDC systems via interoperability protocols. The 2022 prototype demonstrated that central banks could retain sovereign control of their CBDC systems while enabling cross-border settlement through standardized connectivity.

The interoperability model from Dunbar has influenced how the ECB is thinking about digital euro’s cross-border dimension: rather than requiring a global shared ledger (the mBridge approach), the ECB has stated it prefers an “interlinking model based on common standards” for cross-currency digital euro transactions. This matters for operators because it means the digital euro’s cross-border architecture will look more like Project Nexus (the ASEAN payment linkage model) than mBridge.

e-CNY: What Production at Scale Looks Like

China’s e-CNY is the most operationally advanced retail CBDC, giving the rest of the world a preview of what government-issued digital currency looks like in a real economy. The 16.7 trillion yuan ($2.3 trillion) in cumulative volume through November 2025 is real, though context matters: this is cumulative since 2020 trials, and a significant portion reflects government-subsidized distribution (digital red packets at holidays, wage payments, government service payments). Organic merchant adoption has been slower than the volume numbers suggest. The PBoC also announced in December 2025 that e-CNY will add deposit interest features from January 2026, transitioning it from a pure cash-equivalent toward digital deposit money (PBoC, December 2025).

The design choices China made in e-CNY have influenced global CBDC thinking:

Tiered anonymity: e-CNY uses a tiered anonymity model. Small-value transactions (below ~¥2,000 cumulative threshold) are “controllably anonymous” — the PBoC can de-anonymize transaction histories for law enforcement purposes, but day-to-day transactions don’t expose user identity to merchants. Higher-value accounts require identity verification. This two-layer approach is being studied by the ECB for the digital euro.

Offline capability: e-CNY supports offline payments between phones via NFC, without internet connectivity. The phones exchange signed transaction records that are settled when connectivity is restored. This was a specific design requirement for rural China with spotty connectivity — but it’s also relevant for disaster scenarios and the unbanked population globally.

Hardware wallets: China deployed physical smart card hardware wallets for e-CNY, targeting elderly users unfamiliar with smartphones. The dual-form-factor approach (app + card) has been noted by other central banks as a necessary accommodation for demographic inclusion.

Programmability limits: Despite industry speculation about programmable e-CNY (smart-contract-like rules on how CBDC can be spent), the PBoC has been deliberately conservative here. e-CNY is not natively programmable; government subsidy payments have expiry dates implemented at the middleware layer, not embedded in the CBDC token itself. This reflects the PBoC’s view that programmability creates financial stability risks that outweigh the use-case benefits.

Digital Euro: Where It Is and What Operators Need to Model

The ECB completed the digital euro “preparation phase” in 2024 and is advancing through a parallel legislative track (the EU Council and Parliament must pass the Digital Euro Regulation before issuance can begin). The earliest realistic issuance date is 2027–2028, with 2026 occupied by regulation finalization and technical infrastructure tendering.

What is known about the digital euro’s design:

Holding limits: The ECB has indicated a cap of approximately €3,000 per individual will apply to digital euro balances held at the ECB level. This is designed to prevent bank disintermediation — if digital euros are too easy to hold in large quantities, depositors might shift savings from commercial banks to digital euro accounts, destabilizing bank funding. The €3,000 cap is not about transaction limits (which would be higher) but about stored balance.

Merchant acceptance mandate: The digital euro regulation includes mandatory acceptance provisions — payment gateway providers and PSPs operating in the eurozone will be required to accept digital euros from consumers. This is the most directly impactful provision for payment operators: it means accepting digital euro will not be optional once it launches.

Cost to merchants: The ECB has committed that merchants will not pay fees for accepting digital euro in consumer transactions, analogous to the “no charge” model for cash. The cost structure for operators will involve infrastructure integration (comparable to implementing a new payment method) but not per-transaction fees at the scheme level.

Privacy architecture: The digital euro will use a privacy-by-design approach: the ECB will not see individual transaction details; that data is retained by the intermediary (bank or payment service provider) handling the transaction. The architecture is similar to how the ECB today processes card payments — it sees aggregate settlement data, not individual cardholder transactions.

What Operators Should Actually Do

Near-term (2025-2026): No retail CBDC in the EU, US, or most Western markets requires operator action yet. Monitor the digital euro legislative timeline and plan for a mandatory acceptance requirement arriving 2027–2028. Build digital euro into your payment method roadmap as a non-optional integration.

Cross-border operators with Asia and Gulf exposure: mBridge-enabled corridors will become accessible through participating banks in 2026–2027. If you route significant volume through Hong Kong, Thailand, UAE, or Saudi Arabia, begin conversations with your banking partners about their mBridge connectivity timeline. Early-mover operators in these corridors could access lower correspondent banking costs before the market reprices.

Stablecoin positioning: CBDC and stablecoin adoption are not mutually exclusive, but they are competing for the same use cases in cross-border settlement. mBridge settling USD-to-AED in seconds reduces the case for stablecoin rails in that corridor. Monitor which corridors go live on CBDC rails and adjust stablecoin routing strategies accordingly.

Programmability skepticism: The PBoC’s conservative stance on e-CNY programmability reflects a broader central bank consensus that smart-contract-native CBDC creates risks that aren’t justified by available use cases. Operators building products around “programmable money” use cases should not assume CBDC will enable these features — stablecoin platforms are a more realistic substrate for programmable payment logic in the near term.

The CBDC development landscape is moving faster than most operators’ planning cycles. The operators who will be positioned correctly when digital euro launches won’t be the ones who started planning six months before — they’ll be the ones who built the regulatory monitoring and technical assessment process now.