Aani’s 12.5 million users, sarie’s 784 million transactions and Buna’s 100 institutions: what the GCC built first, and why 2027 slipped
By: Christian Brooks – AsiaPresswire – In Riyadh, two payment systems share a name that differs only in capitalisation, and the confusion they generate is not a translation problem. SARIE, in capitals, is the real-time gross settlement system: the wholesale rail on which banks settle with each other, running above SR200 billion a day. sarie, in lower case, is the instant retail scheme that most people mean when they say Saudi Arabia has instant payments. The first settles obligations between institutions. The second settles obligations between people.
The distinction matters because almost every published claim about Gulf payments depends on which one is being counted, and the two numbers differ by roughly an order of magnitude. Saudi Arabia’s instant transfers reached SR7 trillion in June 2026, up 26.1% year on year, but SR5 trillion of that was interbank and SR1.6 trillion was customer payments. Quote the first as evidence of consumer adoption and you are wrong by a factor of three, in a direction that flatters the story.
What is usually called the Gulf’s payments revolution is really three separate projects that happened to run in the same decade: wholesale settlement modernisation, domestic retail instant rails, and a regional cross-border layer. They have different economics, different owners and different success criteria. Only one of them has clearly worked, and it is not the one that gets the conferences.
Two systems with one name in Riyadh
The wholesale side is the older story and the less interesting one. Gulf central banks have operated RTGS systems since the 1990s, and what has changed since is throughput, message format and hours of operation rather than the underlying function. SARIE’s Q1 2026 volume of SR17.78 trillion, up 16.1%, is a strong number that describes an economy settling its interbank obligations faster, not a population changing how it pays for coffee.
The retail side is where the institutional choices get interesting. Saudi Arabia’s instant scheme cleared 784 million transactions in 2025, up 32%, and now runs above 100 million transactions a month on average. The Mada card network, which most foreigners have never heard of, handles more than 8 billion transactions a year across roughly 1.4 million point-of-sale terminals.
Put those together and the 85% figure becomes legible. Electronic payments were 85% of retail payments by value in 2025, against 79% in 2024 and 70% in 2023, the level Vision 2030 had set as a target and which was reached two years early. Cash withdrawals fell 5.4% to SR39.38 billion over the same period. That is a genuine shift in behaviour, and it took a card network, a mandated salary system, a national QR standard and a decade of merchant terminal subsidies to produce.
What the 85% share counts
A scholar’s instinct here is to ask what the denominator is, because the 85% number is doing more work than it can bear. Retail payments by value in Saudi Arabia include a great deal of high-value commerce: property transfers, vehicle purchases, corporate procurement settled by card. A country that digitises its large transactions will move this ratio quickly without changing the payment habits of a household buying vegetables.
The measure that would be more informative is the share of transactions by count, and the share of the population that uses electronic payment weekly. Neither is published in the same release. This is not a Gulf peculiarity; every central bank publishes the ratio that looks best, and the FSB has spent five years asking for a common set of indicators precisely because national numbers are not comparable.
What can be said with confidence is that Saudi Arabia’s cash withdrawal decline is real and large, and that a retail rail carrying a billion transactions a quarter has passed the point where it can be dismissed as a novelty. The disagreement is about magnitude, not direction, and magnitude is where policy arguments actually happen.
Aani was built by a central bank, not by a super-app
Compare the two Gulf approaches and a structural difference appears immediately. Saudi Arabia’s instant scheme was built inside the existing banking system, on top of a card network and a national switch that already had reach. The UAE’s Aani was built as new infrastructure, by a subsidiary of the central bank, with a mandate that every licensed bank join.
The mandate is the interesting part. Aani is operated by Al Etihad Payments, wholly owned by the Central Bank of the UAE, and it launched in October 2023 with compulsory participation for licensed financial institutions. That design solves the two-sided market problem by fiat: there is no chicken-and-egg when the regulator can require the chicken to show up. Seventy-four licensed institutions are now connected, which the operator describes as about 85% of banks, 10% of exchange houses and 5% of digital wallets and finance companies.
Compare that with Southeast Asia, where the successful instant rails were built around super-apps and then opened to banks, or with India, where UPI was built by a non-profit coalition of banks under central bank direction. The Gulf variant is the most centralised of the three, and it has produced the fastest coverage growth: from 1.5 million registered users in early 2025 to 12.5 million by April 2026, with about 774,000 merchants accepting it and settlement in under three seconds.
The numbers, side by side
| Feature | Saudi Arabia | UAE |
| Retail instant rail | sarie / IPS, inside the banking system | Aani, operated by a central bank subsidiary |
| Launch | Incremental build on existing switch | October 2023, mandatory participation |
| Scale, 2025–26 | 784 million transactions in 2025, +32% | 12.5 million users, 74 institutions, 774,000 merchants |
| Card scheme | Mada, 8 billion transactions a year, 1.4 million terminals | Jaywan, live issuance from July 2026 |
| Merchant discount rate | Regulated through the national scheme | 1.5–2.8% credit, 0.8–1.5% debit, no cap |
| Cross-border layer | Buna, AFAQ | Buna, AFAQ, Aani-UPI link agreed |
The table also shows why the two markets are not really comparable. Saudi Arabia’s digital payment story is a card story that added an instant rail. The UAE’s is an instant rail that is now adding a card scheme, because Jaywan only began issuing cards in July 2026 with two banks.
Sixfold growth and a 0.2% activation rate
Both of the following are true of Aani, and the fact that they are both true is the most useful thing about the UAE’s data. Transaction volumes grew sixfold year on year. And the daily activation rate, meaning the share of registered users who transact on a given day, was reported at 0.2%.
Sixfold growth off a small base and a fraction of a per cent of daily engagement are not contradictory. They describe a rail that has been adopted as a registration and not yet as a habit. Twelve and a half million registrations is also, on its face, more than the UAE’s resident population, which is usually estimated at around eleven million. That does not mean the number is false; it means it counts registrations, and a person with accounts at three banks can be three registrations. No payment system in the world reports unique users when it can report registrations instead.
The 0.2% figure is the one a supervisor should care about, because it measures the thing that determines whether the infrastructure pays for itself. A rail with twelve million registrations and a 0.2% daily active rate is processing far fewer transactions than its capacity, which means the per-transaction cost of running it is higher than the business case assumed, which in turn means either the operator absorbs the difference or the pricing changes.
Buna, AFAQ and the layer beneath them
Underneath the domestic rails sits a regional layer that rarely appears in national announcements and matters more than it is credited for. AFAQ, operated by the Gulf Payments Company, connects the RTGS systems of the GCC states so that a payment from a UAE company to a Saudi supplier can clear without routing through a correspondent bank in New York. Buna, owned by the Arab Monetary Fund and headquartered in Abu Dhabi, connects more than 100 financial institutions across Arab jurisdictions and settles in six currencies.
The economics of both are straightforward and unglamorous. A cross-border payment through correspondent banking involves a chain of institutions, each taking a fee and a foreign exchange spread, and each adding a day. A regional settlement system collapses the chain into one hop. The saving is not in the technology, which is ordinary messaging; it is in the number of balance sheets that have to be touched.
Their limit is equally straightforward. AFAQ settles between central banks, so it serves the banks that participate in those systems, which is to say large institutions. Buna serves its members, which is to say banks. Neither was designed to move a hundred dirhams from a warehouse worker to his family in Kerala. That job still belongs to exchange houses and to remittance companies, which is why the UAE’s retail cross-border flows move through a channel that the new infrastructure has barely touched.
What is usually called cross-border reform is domestic plumbing
The G20 roadmap, endorsed in 2020, set three targets for 2027: three-quarters of cross-border retail payments settling within an hour, a cost below 1% for retail and 3% for remittances, and pricing transparency as the norm. The Financial Stability Board’s own progress report, as of late 2025, found just over a third of retail cross-border payments settling within an hour, and nearly one in five remittance corridors still above the 3% ceiling. The FSB acknowledged that the targets were unlikely to be met on the original timetable.
The reason is not a lack of technology, and it is not a lack of ambition. It is that cross-border payments are a coordination problem among jurisdictions whose domestic systems are at different stages. A corridor is only as fast as its slower leg, and its cost is the sum of every intermediary that has to be satisfied plus the foreign exchange spread, which is set by whoever holds the currency at the moment it is needed.
The Gulf’s contribution to this problem has been to fix the domestic leg first and then extend outward, rather than waiting for a global standard. That is why Aani could be linked to India’s UPI in an agreement signed in October 2024, and why the UAE-India corridor, which carries the largest share of the UAE’s outward remittances, is the obvious first test. It is also why the GCC’s own corridor integration through AFAQ has progressed faster than the global roadmap: six central banks with pegged currencies and a shared legal tradition are an easier coordination problem than two hundred jurisdictions.
One third settle within an hour
The one-third figure deserves to be read against the direction of travel rather than as a failure, because the composition of cross-border payments is changing faster than the infrastructure is. A growing share of what crosses borders is not a family remittance but a business payment: a marketplace paying sellers in eight countries, a trading company settling with a supplier, a platform distributing earnings to creators.
Those payments behave differently from remittances. They are larger, they are recurring, they have an invoice attached, and the payer will accept a day of settlement in exchange for a documented trail. They are also the payments that a domestic instant rail can serve, if the last mile is a local account credit rather than an international wire. This is the model that the UAE’s non-oil trade profile makes inevitable: Dh3 trillion of non-oil trade in 2024, Dh1.7 trillion in the first half of 2025 alone, growing at a rate the WTO was forecasting the rest of the world would not match.
The distinction between a remittance and a business payment also explains why the cost targets were written the way they were. A 1% ceiling is achievable for a business payment because the amount is large enough for the fixed costs to disappear into the spread. A 3% ceiling for remittances is difficult for the opposite reason: the average ticket is small, the corridor is thin, and every intermediary has to be paid something. The same infrastructure serves both, but the economics do not converge, which is why the G20 wrote two numbers instead of one.
MDR without a cap is a distributional choice
The UAE has no government cap on merchant discount rates. Credit card acceptance costs a merchant 1.5% to 2.8%, debit 0.8% to 1.5%. Compare India, where UPI merchant payments carry zero MDR by mandate, or China, where acquiring fees sit under a government-guided schedule.
This is a policy choice with a clear distributional consequence, and it is rarely described as one. A merchant accepting a card at 2% is paying a tax on the transaction that goes to the acquiring bank, the issuer and the scheme. For a supermarket with 3% net margins, that is most of the profit on the sale. For a government collecting fees through Aani, the marginal cost is close to zero, which is precisely why the Cabinet authorised federal ministries to accept Aani and Jaywan for fees and fines in 2026.
The absence of a cap is defensible: it keeps the acquiring market commercially viable, it funds fraud protection and chargebacks, and it avoids the Indian outcome where zero-MDR has made card acceptance uneconomic for small merchants and pushed them onto a rail that banks lose money operating. But it also means the cost of digital acceptance is borne by the merchant rather than by the payer or the taxpayer, and in a market where small merchants are the majority of the 774,000 Aani acceptors, that is a choice worth stating out loud.
Jaywan and the sovereignty argument
Jaywan is the UAE’s domestic card scheme, established in 2024 and operated by the same central bank subsidiary as Aani. It began live card issuance in July 2026 with First Abu Dhabi Bank and Commercial Bank of Dubai, after co-badging agreements signed in early 2025 with Visa, Mastercard, UnionPay and Discover.
The sovereignty argument for a domestic scheme is usually made in terms of national pride, which is the least interesting version of it. The operational version is about switching fees and data. Every domestic transaction processed by an international scheme pays a switching fee and generates data that sits outside the country. At the scale of the UAE’s domestic card volume, those fees are a real annual outflow, and the data has regulatory value that is hard to price.
The strategy of co-badging rather than replacement is the pragmatic version, and it is the same one India tried with RuPay and China with UnionPay before restricting foreign schemes. Whether Jaywan reaches the coverage that makes the switching economics work depends on issuance, and issuance depends on merchants, and merchants depend on consumers carrying the card. That loop takes years, which is why the scheme’s real test is 2029 rather than 2026.
ISO 20022 and a migration in two stages
There is one change in this stack that will matter more in five years than any of the consumer-facing launches, and it has no ribbon attached. The global migration of cross-border payments to the ISO 20022 messaging standard runs in two stages: the coexistence period opened on 22 November 2025, and the older MT message formats are retired for cross-border payments on 14 November 2026.
What the standard changes is the amount of information that can travel with a payment. The old formats carried a remitter, a beneficiary, an amount and a small amount of free text. The new ones carry a structured postal address, a legal entity identifier, a purpose code, an invoice reference and enough detail for a compliance system to decide automatically whether a payment is in scope for a sanction or a tax rule.
For a corporate treasurer in Dubai paying a supplier in Riyadh, the practical effect is that the payment stops being an opaque instruction that a bank has to interpret and becomes a data record that a machine can route. The saving is not in the fee; it is in the manual handling that disappears. For a bank, the cost is a multi-year programme that touches every core system, which is why the smaller institutions in the region are the ones most likely to be late.
The Gulf’s advantage here is that several of its newer systems were built as ISO 20022-native rather than retrofitted. SARIE, UAEFTS and Buna were designed or rebuilt in the new format, which means the regional layer does not have to translate between an old domestic standard and a new international one. That is a quiet, unglamorous benefit of having built late, and it will show up in straight-through processing rates long before it shows up in a press release.
The distinction that survives the next cycle
The Gulf has spent a decade building payment infrastructure in the correct order: wholesale settlement, then domestic retail, then the regional layer, then the international interface. That sequencing is why the region’s numbers look better than the global average, and why the FSB’s stalled roadmap does not describe what has happened in Abu Dhabi or Riyadh.
But the sequencing also creates a specific blind spot. Every layer built so far serves institutions or serves consumers at the point of sale. None of it serves the small business that needs to pay a supplier in another currency, or the migrant worker whose remittance still costs more than 3%. Those flows moved through exchange houses twenty years ago and they move through exchange houses today, with a thin layer of fintech on top.
The distinction worth carrying forward is not between fast and slow payments. It is between payments that a central bank can mandate and payments that require two parties in different jurisdictions to agree. The first category is nearly solved in the Gulf and largely unsolved almost everywhere else, which is why the region’s model is being copied. The second is where the remaining cost sits, and no amount of domestic infrastructure will move it.