Major US banks are already answering that question by restricting outside agents. Their concern is understandable. An assistant using a customer’s login can look like the customer while making decisions the customer never specifically approved.

But banks also have a commercial interest in controlling that access. An agent that manages money across institutions could make deposits much easier to move and weaken the bank’s position as the customer’s main financial interface.

The access rules being written now will determine how much of that competition can develop.

What banks are actually blocking

JPMorgan’s Jamie Dimon warned about “rogue agents” during the bank’s October earnings discussion. Reuters subsequently reported that US banks were blocking third-party agents, including transactions flagged after customers shared credentials. Morgan Stanley said it had not identified rogue activity on its platforms, an important qualification to the wider reporting.

Citi’s position draws a useful boundary: outside agents cannot log into customer accounts or transact directly, while read-only access through certain open-banking providers remains available. Bank of America has also imposed restrictions. Brian Moynihan has called for human involvement in the process.

“Banks are blocking AI” therefore misses the detail that matters to product teams. Reading a balance through an approved connection and operating a banking session with someone’s password are different access models. Neither should automatically confer permission to move money.

Customers may welcome an assistant reviewing spending. That does not tell the bank whether it can trust the same assistant to change a beneficiary or submit a transfer.

A successful login leaves the hardest questions unanswered

Consider a customer asking an agent to “put my spare cash somewhere better.”

The customer probably expects the agent to leave enough for rent and upcoming bills. They may assume the destination has the same protections as their existing account. They may never have considered whether a transfer takes several days to reverse.

Read more

In this episode of the WRAP UP podcast, I sit down with Eynat Guez, CEO and co-founder of Papaya Global, to discuss how artificial intelligence and stablecoins are reshaping global employment, payroll and cross-border payments.

Subscribe now

Papaya Global has evolved from a workforce management platform into a global payments business processing close to $50 billion annually, according to Eynat.

She shares how Papaya has reduced its human marketing team by roughly half while tripling its use of AI agents. She also predicts that employees will soon manage multiple AI agents and that governments could eventually introduce taxes based on AI token consumption.

We explore why Eynat still refuses to give AI agents direct access to payment systems, despite her confidence in the technology.

The conversation then turns to stablecoins. Eynat explains why Papaya is already using stablecoin rails to move money internationally and predicts that demand for stablecoin payments could reach 30–40% within two years.

In this episode, we discuss:

• How AI agents could reshape employment and hiring by 2030.

• Why Papaya Global reduced its marketing headcount while increasing AI adoption.

• Whether governments will start taxing AI agents.

• Why AI agents still cannot be fully trusted to execute payments.

• How stablecoins are changing global payroll and treasury operations.

• Papaya Global’s Banco offering and its vision of becoming a new bank for the workforce.

• How Papaya’s partnership with eToro brings savings and investment options to workers.

• Why the future workforce could involve people working for five or ten companies simultaneously.

LinkedIn: https://www.linkedin.com/in/sirojboboev/

Instagram: https://www.instagram.com/wrapuppodcast/

Tiktok: https://www.tiktok.com/@wrapuppodcast

X: https://x.com/samboboev

European banks and fintechs expect payment rules to drive their biggest operational changes over the next two years, while KYC remains their largest source of pressure today.

In its State of European Financial Services 2026 report, Sumsub asked which compliance areas created the greatest operational pressure over the past 12 months.

KYC led at 32%. Sanctions screening followed at 27%, with AML transaction monitoring at 26% and payment processing requirements at 25%.

The small gaps matter. Firms are dealing with several demanding control areas at once. Business verification also stood at 25%, while data privacy reached 24%.

There are regional differences. Sanctions screening takes the top spot in Central Europe, Eastern Europe and the Caucasus. But the same core pressures keep appearing across Europe, even where the leading fraud threats differ.

Last year, 51% of respondents identified keeping up with changing regulations as a major compliance challenge. This year’s survey gives a clearer picture of where that pressure lands inside the business.

Sumsub then asked which regulatory developments would require the greatest operational change over the next 24 months.

The Instant Payments Regulation led at 27%, closely followed by Verification of Payee at 26%. One or both also led expectations in the UK and Ireland, Iberia, DACH and the Baltics.

Payment changes have become a shared concern across markets.

The EU AI Act followed at 22%. DORA and the EU AML Regulation or Sixth AML Directive each stood at 20%. PSD3 or the Payment Services Regulation reached 18%, alongside AMLA.

For banks and fintechs, this means payment delivery plans need to account for compliance work already underway.

Verification of Payee brings decisions directly into the customer journey. When a name does not match, the firm needs to explain the result clearly and help the customer decide what to do next. That takes product design and operational support as well as a technical integration.

My reading of the findings is that execution capacity will become a bigger constraint. Several programmes can have separate budgets and still depend on the same engineers.

A payment deadline does not clear the KYC backlog. Moving those engineers onto a new regulatory programme can leave existing control improvements waiting for the next release.

READ FULL REPORT

This week’s fintech landscape highlights two major shifts: the rise of AI-powered commerce and the growing role of stablecoins in payments. AI agents are moving beyond product discovery toward autonomous purchasing and financial execution, but consumer concerns around privacy, accuracy, fraud, and accountability remain significant. Meanwhile, stablecoins are gaining traction in cross-border payments, business transactions, and yield-generating products, although regulatory requirements and conversion costs still limit adoption. DeFi is also reshaping financial infrastructure through automated lending, trading, and settlement. According to BCG, global payments revenue could reach $2.6 trillion by 2030, with growth increasingly driven by emerging markets, while rising operating costs pressure profitability. The key takeaway: the next phase of fintech competition will be about more than faster payments; it will depend on earning consumer trust, building reliable AI-driven infrastructure, and turning payment capabilities into sustainable revenue.

Video of the Week

Deep Dive of the Week

Ant International’s Agentic Mobile Protocol

Software agents are transitioning from conversational text generators into autonomous economic actors capable of executing commercial workflows. Recent Global Payments research found that Americans expect AI agents to make 15% of their purchases within five years, up from 9% a year earlier. Research from 451 Research shows that 58% of organizations are actively seeking opportunities to implement AI agent capabilities. By 2030, AI agents are projected to orchestrate $3 trillion to $5 trillion of global consumer commerce.

However, an execution gap remains between using AI for discovery and allowing it to act. Research from Alipay+ and S&P Global found that 81.3% of surveyed consumers already use AI-powered tools to explore travel destinations and experiences. Interest falls as AI moves closer to payments and autonomous decision-making, with 26% expressing interest in those capabilities. The gap points to a trust problem that payment providers will need to address before autonomous execution becomes mainstream.

This week’s reports

1️⃣How much do bank customers trust gen AI?

2️⃣The Enterprise Guide to Stablecoin Earn Programs

3️⃣Stablecoins and World Trade

4️⃣Decentralized Finance (DeFi): A new infrastructure for investment funds

5️⃣A short history of the future of shopping and payments

6️⃣The Agentic Commerce Report

7️⃣The Burden of Proof

How much do bank customers trust gen AI?

Many banking customers use gen AI tools to research finances but hesitate to share personal data, trust recommendations, or let AI act on their behalf. What can that mean for banks?

Banking consumers are using more generative AI1 as a way to search, sort, and compare options, but Deloitte survey data suggests they remain selective about where they’ll allow AI to go next because they are worried about mistakes and privacy lapses.

Deloitte surveyed almost 2,600 banking customers in the United States and found that 72% of consumers across different ages, genders, and income levels are concerned about sharing information about their financial situation with gen AI tools. Many (64%) also worry about hidden biases. Only 46% of respondents trust the accuracy of banking recommendations from the tools.

By contrast, 79% of bank customers who responded to the survey trust information from their banks’ own websites when researching banking products. That compares with consumers’ trust in traditional search engines (67%), gen AI tools (49%), and social media (28%) (figure 2). This difference between bank websites and gen AI tools reflects trust built by banks over years of regulations and customer experiences. Institutions should sustain this influence by ensuring their information and products are visible as gen AI tools become more popular gateways to financial research.

Read more

Software agents are transitioning from conversational text generators into autonomous economic actors capable of executing commercial workflows. Recent Global Payments research found that Americans expect AI agents to make 15% of their purchases within five years, up from 9% a year earlier. Research from 451 Research shows that 58% of organizations are actively seeking opportunities to implement AI agent capabilities. By 2030, AI agents are projected to orchestrate $3 trillion to $5 trillion of global consumer commerce.

However, an execution gap remains between using AI for discovery and allowing it to act. Research from Alipay+ and S&P Global found that 81.3% of surveyed consumers already use AI-powered tools to explore travel destinations and experiences. Interest falls as AI moves closer to payments and autonomous decision-making, with 26% expressing interest in those capabilities. The gap points to a trust problem that payment providers will need to address before autonomous execution becomes mainstream.

Legacy payment rails exacerbate this gap. Traditional card-not-present infrastructure relies on human interaction points such as web redirects, visual multi-factor authentication screens, and manual passcodes. Automated agents cannot complete visual verification challenges or enter one-time passwords without custom workarounds. Existing fraud detection systems regularly flag programmatic agent requests as malicious bot traffic, leading to transaction rejections. Traditional card economics can also make high-frequency, sub-cent transactions impractical.

To address these limitations, payment providers are developing agent-native execution layers. Ant International launched the open-sourced Agentic Mobile Protocol (AMP) in April 2026. In its first rollout phase, integration is supported by 10 Alipay+ wallet partners - which together represent 1.5 billion user accounts - and eight leading acquirers: Adyen, Allinpay, Antom, Checkout.com, Fiserv, Global Payments, Nuvei and Worldline. Deploying agentic capabilities across an ecosystem of this size requires an abstraction layer that maintains mobile wallet security while offering programmatic transaction routing.

Enterprise purchasing offers an immediate growth vector alongside consumer applications. In business-to-business environments, agents can manage software subscriptions, media ad spend, and procurement within predefined policy limits. Speaking at the Voyage conference, Ant International Chief Innovation Officer Jiang-Ming Yang noted that he expects adoption to progress faster in business-to-business contexts than in consumer markets in the near-term. Business procurement workflows follow structured operational steps, strict compliance rules, and measurable objectives. This structure simplifies return on investment calculations and accelerates corporate deployment.

Agentic Mobile Protocol architecture

AMP addresses execution friction by connecting AI agents to existing mobile payment infrastructure while keeping authorization boundaries around the user’s account. Its public architecture centers on Know-Your-Agent controls, permissioned delegation, wallet connectivity and agent-to-agent settlement. For payment teams, these capabilities can be understood across three functional layers: trust and identity, payment execution and wallet authorization.

The trust layer starts with Know-Your-Agent (KYA). Ant International describes KYA as a verification process that establishes an agent’s digital identity and certifies its authorized capabilities. AMP adds an Agent Trust Rating, a dynamic risk-management mechanism that assesses whether an agent can be trusted and controls the level of autonomy it receives. Together, these controls are designed to establish who or what is initiating a transaction and what the agent has permission to do.

Industry convergence on identity standards is critical for cross-border scale. Ant International, Visa and Mastercard have begun collaborating on a KYA interoperability framework. The three companies have already developed their own approaches through Ant International’s AMP, Visa’s Trusted Agent Protocol and Mastercard’s Verifiable Intent. The collaboration will explore common principles for agent identification and onboarding across networks while preserving each network’s own verification and decisioning processes. The work also connects with BuildFin.ai, an industry platform convened by the Monetary Authority of Singapore. The collaboration builds on the Safeguards for Agentic Finance at Runtime, or SAFR, framework, which is being developed around safeguards for AI agents operating in financial services.

The Payment Service Layer converts validated agent requests into standardized payment instructions without exposing credit card details, bank credentials, or private API keys. Developers integrate this layer into AI models using open-source software development kits without modifying merchant checkout flows. According to Ant International, AMP cuts the number of steps required to link a payment agent to a digital wallet by 50% compared with traditional card-binding methods.

The Wallet Service Layer maintains mobile wallets at the core of authorization. Mobile payment apps retain administrative control over account funding, multi-factor authentication triggers, velocity limits, and daily spending caps. Before completing the payment, the wallet infrastructure verifies that agent requests match user-defined authorization rules, giving account holders continuous control over financial assets.

How to approach autonomous liability

Assigning transaction liability remains a central challenge in agentic commerce. When an agent executes an incorrect purchase or experiences an operational error, payment networks require clear rules to determine financial responsibility. Global Payments research indicates that 50% of consumers cite payment security concerns, while 33% demand explicit per-transaction approval before payments process. The Agentic Mobile Protocol addresses these concerns by establishing two interaction models backed by system risk guarantees.

Immediate journeys cover transactions where the user remains involved during checkout. The agent can handle discovery, product selection and preparation of the transaction while the wallet retains the final authorization step. This reduces manual interaction without removing user control over payment.

Delegated journeys extend the model to tasks that users authorize in advance. AMP is designed around clear permission boundaries: the user authorizes a task rather than handing the agent unrestricted access to an account. Ant says users retain real-time visibility and can modify or revoke delegated tasks. This model allows an agent to operate within an approved mandate while keeping authorization boundaries around its actions.

AMP also introduces AgentSafePay as a risk layer for agent-initiated transactions. Ant International describes it as providing a money-back guarantee for merchants against agentic-specific risks.

The protocol also includes a high-frequency agent-to-agent settlement engine designed for micro-transactions. Traditional card network fee structures impose fixed costs that make micro-payments inefficient. Ant International says AMP’s high-frequency A2A settlement mechanism can handle transactions as small as $0.000001 between agents, with real-time accounting and clearing. This capability enables software agents to purchase data feeds, compute capacity, and API access on a granular pay-per-use basis without incurring standard card interchange overhead.

Full-stack agentic integration

Global commercial operations require integrated financial technology solutions that extend beyond checkout buttons. As enterprises expand across international markets, operational complexities demand automated management across multi-currency accounts, foreign exchange, treasury operations, and growth workflows. Market research from 451 Research highlights that 58% of enterprises actively seek automated systems to manage cash flow, fraud detection, and credit risk.

Within this operational stack, multi-currency accounts automate cross-border clearing and liquidity management. Foreign exchange management utilizes specialized AI models to manage currency volatility. Ant International deploys FalconTST for FX and cash-flow forecasting. The company says the model has more than 8.5 billion parameters and has achieved more than 93% accuracy in real-world FX and cash-flow forecasting cases. In those deployments, Ant reports that FalconTST reduced corporate FX hedging and allocation costs by between 30% and 60%.

Growth operations and merchant workflows similarly deploy specialized software agents. Antom Autopilot handles the entire merchant payment lifecycle through a single conversational window, from onboarding and risk protection to revenue recovery and new market expansion. Antom also applies AI to time-intensive merchant tasks like dispute management. Its Antom SmartDispute advises on full dispute management and appeal strategies, and can drive average dispute success rate from 24.5% to 41.8%. Underlying these services is a transformer model that Ant International calls its 3-in-1 payment foundation model. The model combines graph, sequential and tabular data, and identifies high-risk transactions with more than 95% precision and has increased payment success rates by up to 13.5%. Ant International notes that the foundation model is adaptable to general purpose tasks across a range of payment scenarios, from payment protection to account security. The company highlighted that it has now replaced over 200 legacy point solutions with this 3-in-1 transformer model, which allows it to avoid rebuilding from zero when new threats emerge, cutting model development time by 80%.

Ant International says the number of merchants using AI for payment integration quadrupled over the past twelve months. Under Antom, 89.5% of merchants have deployed AI agents, while 81.4% of payment tasks are aided by AI. The company has not disclosed the underlying merchant count or methodology behind these percentages, so the figures are best treated as company-reported indicators of adoption rather than market-wide benchmarks.

As software agents take on more commercial activity, payment infrastructure has to solve a different set of problems from today’s checkout stack. The agent needs an identity. Its authority needs boundaries. Merchants and payment providers need to know who stands behind it. Payment networks also need a way to carry those trust signals across different rails.

AMP is Ant International’s attempt to build that layer around mobile payments. Its scale will depend partly on adoption across Alipay+, but the more consequential work may be happening outside Ant’s own network. The KYA collaboration with Visa and Mastercard is an early test of whether competing agent-payment protocols can establish common trust principles without requiring merchants to integrate another fragmented set of standards.

Sources

1. Consumers Expect AI to Make 15% of Their Purchases Within Five, https://investors.globalpayments.com/news-events/press-releases/detail/516/consumers-expect-ai-to-make-15-of-their-purchases-within

2. Payment trends update: how agentic AI is disrupting payments, https://www.visaacceptance.com/en-us/blog/article/2025/how-agentic-ai-is-disrupting-payments.html

3. Ant International, Mastercard and Visa Initiate Collaboration on, https://www.ant-intl.com/en/news/detail/?id=ant-international-mastercard-and-visa-initiate-collaboration-on-know-your-agent-interoperability-to-scale-agentic-commerce

4. Ant International Launches Open-Sourced Agentic Mobile Protocol, https://www.ant-intl.com/en/news/detail/?id=ant-international-launches-open-sourced-agentic-mobile-protocol-to-drive-ai-commerce&tag=AI

5. Interoperability and Trust Key to AI Commerce: Alipay+ and S&P, https://www.newswire.co.kr/newsRead.php?no=1041724

6. Ant International’s Agentic Mobile Protocol Rolls Out Globally with, https://www.ant-intl.com/en/news/detail/?id=ant-internationals-agentic-mobile-protocol-rolls-out-globally-with-wallets-acquirers-initiating-collaboration-on-kya-interoperability-framework-with-mastercard-and-visa&redirect=news&tag=AMP%2CPARTNERSHIP

7. Expanding infrastructure for the age of AI commerce - Ant International, https://www.ant-intl.com/en/news/detail/?id=expanding-infrastructure-for-the-age-of-ai-commerce-ant-international-connects-over-150m-merchants-with-more-than-2b-consumers

8. MAS Partners Industry to develop Safeguards for AI Agents in Finance, https://www.mas.gov.sg/news/media-releases/2026/mas-partners-industry-to-develop-safeguards-for-ai-agents-in-finance

9. Ant International Launches Industry’s First Full-Stack AI-Native Solutions for Payment, Account, FX, Treasury and Growth Operations for Global Businesses
   https://www.ant-intl.com/en/news/detail/?id=ant-international-launches-full-stack-ai-native-solutions-for-payment-account-fx-treasury-and-growth-operations-for-global-businesses&redirect=news&tag=VOYAGE%2CAI%2CANTOM%2CALIPAY%2B%2CPARTNERSHIP%2CFALCONTST%2CBETTR

Disclaimer:

Fintech Wrap Up aggregates publicly available information for informational purposes only. Portions of the content may be reproduced verbatim from the original source, and full credit is provided with a “Source: [Name]” attribution. All copyrights and trademarks remain the property of their respective owners. Fintech Wrap Up does not guarantee the accuracy, completeness, or reliability of the aggregated content; these are the responsibility of the original source providers. Links to the original sources may not always be included. We use AI tools to brainstorm, draft, edit, and summarize, but the ideas, analysis, and connections Sam draws are his. Every piece is verified and written in Sam’s voice. For questions or concerns, please contact us at sam.boboev@fintechwrapup.com.

In this episode of the WRAP UP podcast, I sit down with Jordan Sinclair, President of Robinhood UK, to explore how artificial intelligence, blockchain technology, and changing consumer expectations are reshaping the future of investing.

Subscribe now

Robinhood is expanding beyond traditional stock trading into crypto, tokenized assets, prediction markets, and AI-powered investing. But the wider financial industry is undergoing a similar transformation, with companies like Revolut and Coinbase competing to become the primary financial platform for consumers.

One of the most interesting parts of our conversation was about the role of AI agents in investing.

Robinhood has already introduced agentic trading capabilities in the US, allowing customers to connect AI agents to their investment accounts, allocate funds, and define trading rules.

I asked Jordan a question I’ve been thinking about for some time: If my AI agent can research markets, access my portfolio, and execute trades, why would I still need a trading app?

We also explored the potential of tokenization to change how financial assets are owned and transferred. Today, moving stocks between brokers can take weeks. Blockchain-based infrastructure could make these transfers much faster and enable trading outside traditional market hours.

Another topic we discussed was the growing popularity of prediction markets. As trading volumes increase, where should regulators draw the line between financial instruments and gambling?

In this episode, we discuss:

• AI agents and the future of investing: Will AI eventually manage our portfolios and make trading decisions?

• The future of trading apps: Could AI agents replace the way we interact with financial platforms?

• Tokenized stocks and private markets: How blockchain could enable 24/7 trading, faster settlement, and broader access to investments.

• Prediction markets: Are they legitimate investment products or another form of gambling?

• The financial super app race: How Robinhood, Revolut, and Coinbase are competing for customers.

• AI risks and regulation: Who takes responsibility when AI-driven investment decisions go wrong?

• The future of retail investing: Why millions of people still don’t invest and how technology could change that.

LinkedIn: https://www.linkedin.com/in/sirojboboev/

Instagram: https://www.instagram.com/wrapuppodcast/

Tiktok: https://www.tiktok.com/@wrapuppodcast

X: https://x.com/samboboev

This week’s Fintech Wrap Up explores how AI, stablecoins, and regulatory changes are reshaping financial services. The Evident AI Index reports a 26% increase in banks’ average AI scores, highlighting accelerated adoption but limited disclosure of measurable financial returns. Meanwhile, Europe’s e-invoicing mandates could modernize B2B payments and unlock embedded financing opportunities, while euro stablecoins and stablecoin-linked cards continue expanding digital payment infrastructure. Crypto and traditional finance are also converging through more than $22.8 billion in announced M&A deals since January 2024. Finally, as banks deploy autonomous AI agents, stronger real-time risk monitoring and governance will become essential to managing operational risks.

Video of the Week

Deep Dive of the Week

Personal Agent Protocol and Agentic Commerce

The rapid rise of autonomous artificial intelligence agents acting on behalf of individual consumers has exposed fundamental limitations in traditional web infrastructure. Until recently, consumer AI agents interacted with businesses primarily by emulating human browsing behavior. These software proxies loaded standard web pages, parsed document object models, filled out form fields, and clicked navigation buttons. When web interfaces proved too complex or brittle, agents escalated tasks by dialing customer support call centers or initiating live chat sessions. This browser emulation model introduced severe operational friction. Document object model parsing is fragile, subject to failure whenever a merchant updates its user interface. The resulting interactions suffered from high latency, excessive compute overhead, and high task failure rates.

This week’s reports

1️⃣How Europe’s E-Invoicing Mandates Rewire B2B Payments

2️⃣The state of euro stablecoins in 2026

3️⃣Stablecoin Cards: The Infrastructure Race for the Next $100 Billion

4️⃣AI-powered personalization can help banks create value

5️⃣Evident AI Index for Banks

6️⃣Crypto and TradFi are converging through M&A

7️⃣When Agents Run the Bank

How Europe’s E-Invoicing Mandates Rewire B2B Payments

Those who think of Europe’s B2B e-invoicing mandates as just a tax-compliance software upgrade have clearly gotten it wrong. Beyond the compliance burden, these mandates unlock a larger commercial opportunity: making structured, machine-readable transaction data ubiquitous and cost-effective to access. This data layer is also the missing precondition for two propositions that have been stuck at the margins of European B2B payments for decades: commercial card acceptance on the supplier side, and fintech-led working-capital financing embedded in CFO office platforms.

A Regulatory Reform That Is Quietly a Payments Reform

The European invoice is being digitized, and key regulations are underway. ViDA (VAT in the Digital Age) entered into force with mandatory e-invoicing and near-real-time digital reporting for intra-EU B2B transactions (applicable from July 2030). National B2B e-invoicing mandates are well underway, dematerializing paper invoices and standardizing the e-invoice format and transportation interface. For example, Belgium went live with mandatory domestic B2B structured e-invoicing on 1 January 2026. Poland’s KSeF clearance model captured large taxpayers on 1 February 2026 and essentially everyone else on 1 April 2026. France begins its phased rollout on 1 September 2026.

These regulations standardize both the content and the transport of every B2B invoice simultaneously:

Read more

The rapid rise of autonomous artificial intelligence agents acting on behalf of individual consumers has exposed fundamental limitations in traditional web infrastructure. Until recently, consumer AI agents interacted with businesses primarily by emulating human browsing behavior. These software proxies loaded standard web pages, parsed document object models, filled out form fields, and clicked navigation buttons. When web interfaces proved too complex or brittle, agents escalated tasks by dialing customer support call centers or initiating live chat sessions. This browser emulation model introduced severe operational friction. Document object model parsing is fragile, subject to failure whenever a merchant updates its user interface. The resulting interactions suffered from high latency, excessive compute overhead, and high task failure rates.

For merchants, autonomous agents operating through standard web browsers created an acute visibility crisis. Security systems and web application firewalls could not reliably distinguish a legitimate consumer agent performing a routine balance inquiry from a malicious web scraper, an automated attack script, or a credential stuffing bot. Consequently, businesses faced a binary choice between blocking non human traffic at the edge or exposing customer facing interfaces to unmitigated operational and security risks.

What brought us to open standards

The friction between consumer agents and enterprise infrastructure escalated rapidly in late 2026. Meta launched its personal AI agent, Muse, on September 8, 2026, achieving rapid consumer adoption and topping mobile app store charts within days. Designed to navigate websites, book travel, negotiate bills, and execute retail transactions, Muse operated directly across public merchant interfaces. On September 21, 2026, twelve days after launch, Amazon actively blocked Muse from accessing its retail platform, halting automated shopping tasks. This public collision underscored a core reality: digital commerce platforms cannot support millions of persistent, high-frequency AI agents operating through unauthenticated front-door web browsers. Market momentum accelerated further as everyday assistant startup Instinct saw its valuation quadruple to ten billion dollars, OpenAI launched its task-oriented agent suite Dots, and Hark released software designed to execute household purchasing tasks.

To resolve this systemic standoff, Sierra and Meta introduced the Personal Agent Protocol (PAP) on October 6, 2026. Co-developed alongside an initial coalition of enterprise, payments, and retail leaders, including Genesys, Instinct, Rocket Companies, Shopify, Stripe, and Walmart, the protocol establishes an open communication standard for consumer-to-business agent interactions. The initiative replaces unstructured web scraping with standard authorization, identity signaling, and structured task delegation.

Read more

In this episode of WRAP UP, I sit down with Suresh Balaji, Chief Marketing Officer of Lloyds Banking Group, to discuss how Lloyds is approaching AI, agentic systems, customer trust and the future of banking.

Subscribe now

Lloyds serves around 28 million custome…

Read more

European companies with 50 to 249 employees may be losing around €7 billion a year through avoidable SaaS pricing gaps. Duplicate tools could add another €2 billion.

Spendesk’s analysis of 12 months of transaction data across 2,500 European companies shows how quickly software spend becomes hard to manage.

The AI data is the clearest example.

More than two-thirds of companies that pay for AI use at least two competing products. AI spend grew 340% over the year, even though the number of companies buying AI changed very little.

That suggests the cost is rising inside companies that have already adopted AI. Teams are using more models, licences and usage-based products. Each purchase may look small on its own, but together they create a finance problem.

Traditional SaaS budgets were built around predictable monthly or annual fees. AI spend changes with usage. A product can be cheap at rollout and expensive once it becomes part of everyday work.

Finance teams need a clear view of who owns each tool, whether it is being used, and where similar products overlap. Procurement also needs to be involved earlier, before new subscriptions become embedded across teams.

The AI conversation inside companies is moving from experimentation to cost discipline.

Read Full Report

This week’s Fintech Wrap Up highlights how digital finance is becoming more automated, connected, and AI-driven: payment optimisation and e-invoicing are improving efficiency, bank-account data is strengthening fraud and risk detection, while AI is simultaneously accelerating fraud and cybersecurity threats. At the same time, blockchain and tokenised finance are increasingly focused on interoperability between fragmented networks, while businesses continue expanding AI adoption despite cost constraints and uncertain workforce impacts.

Video of the Week

Deep Dive of the Week

How Stripe’s Product Stack Changed In The Past 5 Years

If you run a software enterprise building generative AI applications today, your CFO spends far more time tracking GPU inference bills than traditional server expenses. Every user prompt hits upstream model providers, consumes tokens, and generates variable infrastructure costs. Meanwhile, your product team balances prepaid credits, usage thresholds, and multi-jurisdictional tax compliance across dozens of countries.

Back in 2021, solving this problem required connecting separate vendors for card processing, meter tracking, tax engines, and API routing. Today, this operational and financial workflow sits within Stripe’s expanded product stack.

This week’s reports

1️⃣Closing the retail payments optimisation gap

2️⃣How Europe’s E-Invoicing Mandates Rewire B2B Payments

3️⃣The Bank Account Footprint

4️⃣State of Fraud Report 2026

5️⃣The Connective Tissue of Digital Finance

6️⃣The state of AI in 2026

7️⃣Frontier AI: A New Era of Cyber Resilience

Closing the retail payments optimisation gap

Payments optimisation is becoming a vital part of modern retail. As retailers work across providers, gateways and processors, failed payments, fraudulent transactions and transaction friction can quickly eat into retailers’ profits if left unchecked, creating revenue leakage that is often difficult to identify and recover.

Key findings:

1. Retailers are more likely to measure optimisation through cost reduction than direct revenue impact

60 percent of respondents measure the success of payments optimisation through payment processing costs, while 37 percent measure revenue directly attributed to payments performance gains. This suggests that cost-related outcomes remain more widely tracked than direct revenue contribution.

2. Payments optimisation is recognised as important, but not always prioritised

70 percent of organisations treat optimisation as a topic of standalone importance, with a quarter assigning it board-level visibility. However, fewer than half describe it as either a top-tier strategic priority or a major operational focus.

3. Resource constraints are limiting further improvement

Read more

If you run a software enterprise building generative AI applications today, your CFO spends far more time tracking GPU inference bills than traditional server expenses. Every user prompt hits upstream model providers, consumes tokens, and generates variable infrastructure costs. Meanwhile, your product team balances prepaid credits, usage thresholds, and multi-jurisdictional tax compliance across dozens of countries.

Back in 2021, solving this problem required connecting separate vendors for card processing, meter tracking, tax engines, and API routing. Today, this operational and financial workflow sits within Stripe’s expanded product stack.

What Stripe Already Was in 2021

Understanding this transition requires establishing an accurate baseline. By 2021, Stripe was far more than a basic payment acceptance gateway. Its core ecosystem included Stripe Billing for subscriptions, Stripe Connect for marketplace payout routing, Stripe Radar for fraud prevention, Stripe Issuing for card creation, and Stripe Treasury for embedded banking services.

Software companies relied on this stack to accept online payments, split revenue between platforms, and manage fixed recurring billings.

However, major operational friction points remained difficult for internet businesses:

• Tax Compliance: Calculating and remitting sales tax, VAT, and GST across thousands of local jurisdictions required complex external software integrations.

• Financial Reconciliation: Matching transactions between internal application ledgers, payment gateways, and third-party bank accounts required heavy manual engineering.

• Merchant of Record Overhead: Digital product creators lacked turnkey options to delegate legal selling liabilities, chargeback risks, and global tax filings.

• High-Throughput Metering: Standard billing systems struggled to ingest and calculate millions of real-time usage events for usage-based software.

• Cross-Border Transfers: Moving money internationally remained bound to slow correspondent banking rails, creating clearing delays and transfer fees.

• Embedded Credit: Platform merchants needed access to working capital without navigating slow traditional bank underwriting channels.

The Acquisition Spree

In the past several years, Stripe closed several targeted acquisitions to expand its operational footprint beyond payment processing.

Early acquisitions targeted tax compliance and back-office reconciliation. In April 2021, Stripe acquired TaxJar to launch Stripe Tax, embedding automated tax calculations directly into checkout flows. In October 2021, Stripe acquired Recko to automate complex ledger matching across internal databases, processors, and bank statements.

Read more

In this episode of the WRAP UP Podcast, I sit down with Philip Bruno, Chief Strategy and Growth Officer at ACI Worldwide, to explore how the payments indus…

Read more

This week’s reports explore how AI and digital-asset technologies are reshaping financial services, work, and payments. Agentic AI is moving from experimentation toward real operational use, creating new opportunities for banks and other financial firms while also raising challenges around governance, fraud, accountability, workforce change, and employee adoption. At the same time, tokenization and distributed-ledger-based payments are advancing gradually, with their success depending less on the technology itself and more on practical benefits, regulation, interoperability, trust, and integration with existing financial infrastructure.

Video of the Week

Deep Dive of the Week

The UK and EU Core Banking Companies Directory

I built this guide because core banking shortlists in the UK and EU are still assembled from analyst quadrants, vendor decks and whoever the systems integrator knows best. I profiled 18 core providers against the same set of criteria: architecture, deployment, product coverage, payment rails, compliance posture, pricing and named UK and EU clients. The market moved a lot this year. Finastra agreed to sell its core banking unit, Deutsche Bank’s Private Bank signed with Thought Machine, and several vendors changed CEOs, so older vendor maps are already out of date.

This week’s reports

1️⃣The State of Autonomous Finance 2026

2️⃣Two futures for jobs in an AI era

3️⃣Benchmark AI Agents x Stablecoins

4️⃣The state of AI in 2026: On the road to ROI

5️⃣Distributed Ledger Technology and the Future of Payments in Canada

6️⃣How to close the agentic adoption gap

7️⃣2026 Report on Tokenization Adoption

The State of Autonomous Finance 2026

Two findings frame the position of financial institutions at mid 2026

The first is commercial. A survey of 522 finance leaders published in June found that 71% already factor a bank’s agentic AI capability into choosing or expanding a banking relationship, rising to nearly eight in ten among mid-market companies. 69% said they would pay their bank for secure, compliant, auditable agentic connectivity. Agentic capability has become a selection criterion for commercial banking relationships. The competitive contest is moving from the account layer to the workflow layer, and the ERP systems, treasury platforms, and fintech applications already embedded in clients’ workflows are positioned to capture that role if banks do not.

Read more

I built this guide because core banking shortlists in the UK and EU are still assembled from analyst quadrants, vendor decks and whoever the systems integrator knows best. I profiled 18 core providers against the same set of criteria: architecture, deployment, product coverage, payment rails, compliance posture, pricing and named UK and EU clients. The market moved a lot this year. Finastra agreed to sell its core banking unit, Deutsche Bank’s Private Bank signed with Thought Machine, and several vendors changed CEOs, so older vendor maps are already out of date.

What this guide covers

I cover 18 vendors that sell core banking or core-adjacent platforms to institutions in the UK and EU, grouped into four categories. For each one I checked official websites, press releases, client announcements and regulator registers, and confirmed ownership and naming as of September 2026. Where a vendor does not publish a data point, I mark it “Not disclosed” instead of inferring it.

Subscribers also get the Excel file with nine tabs: Overview, Architecture and Deployment, Products, Target Clients, Regulatory and Compliance, Payments Connectivity, Clients and Pricing, Regions, and a filterable Master Table with every vendor against every dimension.

The 18 vendors

Cloud-native next-gen cores

Thought Machine: cloud-native Vault Core and Vault Payments, now the default shortlist entry for tier 1 core replacements.

10x Banking: SaaS cloud-native core built for large retail banks and building societies.

Engine by Starling: Starling Bank’s own core and digital banking stack, licensed to other banks as managed SaaS.

Incumbent enterprise cores

Temenos: the largest independent core vendor, selling Transact on-premise, in the cloud or as SaaS.

Finastra: Vista-owned vendor narrowing to lending and payments after agreeing to sell its core unit to Pollen Street Capital.

FIS: US processor selling Modern Banking Platform, Profile and, in DACH, KCORE24.

Fiserv: US account processing giant whose cores have almost no UK or EU footprint.

Infosys Finacle: componentised core and digital banking suite sold as licence or SaaS.

Oracle FLEXCUBE: Oracle’s universal banking core, paired with OCI-only Oracle Banking Cloud Services.

SBS: the former Sopra Banking Software, now owned by 74Software, with a deep UK building society book.

Composable and API-first platforms for fintechs and challengers

Mambu: multi-tenant SaaS lending, deposits and payments engines configured through APIs.

Tuum: modular core, formerly Modularbank, sold module by module or as a full replacement.

Skaleet: EU-hosted SaaS core with a native payments engine for neobanks and embedded finance.

Specialist and regional cores

Avaloq: NEC-owned front-to-back platform for private banks and wealth managers.

ERI: OLYMPIC Banking System, a private banking and wealth core strong in Luxembourg and Switzerland.

Banqsoft: Nordic platform for consumer lending, deposits and asset finance.

Profile Software: Athens-listed vendor of the Finuevo core and Axia wealth platform.

CR2: HPS-owned digital banking and card payments layer that sits on top of a core.

What’s behind the subscribe button

Paid subscribers get a structured profile for each of the 18 vendors covering architecture, deployment and cloud partners, products, target clients, licence status, payment scheme connectivity, security certifications and DORA posture, pricing model, named UK and EU clients, and regional coverage. They also get my notes on 2026 ownership changes and the full Excel file with all nine tabs.

Subscribe to unlock the full vendor profiles and the Excel file.

Read more

In this episode of the WRAP UP podcast, I sit down with Eric Young, CTO of Nubank, to discuss how the company is using AI across engineering, credit, and customer experiences.

Subscribe now

Nubank has reported that its weekly AI token use grew almost 10x since the beginning of …

Read more

This week’s Fintech Wrap Up highlights the rapid convergence of AI, payments, digital assets, and financial infrastructure. Key themes include the growing adoption of agentic AI in banking and treasury, with institutions moving toward autonomous systems that can support decision-making, fraud detection, liquidity management, and policy-based actions. Meanwhile, Amazon’s AgentCore Payments is enabling AI agents to make autonomous payments, while crypto exchanges are evolving into broader financial “super apps” combining trading, yield, payments, custody, and collateral. PayPal’s Q2 results show continued revenue growth, while instant payments are expanding globally and putting pressure on traditional payment models. Across Europe, fragmented banking, payments, and investing ecosystems also highlight the challenge of building a truly unified financial super app.

Video of the Week

Deep Dive of the Week

Anthropic’s In-House Payment Push Is Easier Said Than Done

Fast-growing software platforms eventually hit a financial scale where off-the-shelf payment software creates operational friction. Anthropic saw its annualized revenue run-rate jump from $787 million to nearly $45 billion within five months by May 2026. The company reached a $965 billion valuation during its Series H funding round. When transaction volumes grow this quickly, small pricing inefficiencies and billing delays multiply into major operational expenses. Anthropic recently posted job listings for billing software engineers, fraud specialists, and treasury managers to evaluate developing proprietary financial tools.

Industry commentary quickly framed these job listings as a direct threat to Stripe. The full operational context presents a more nuanced reality. Anthropic is evaluating which specific financial primitives to build internally, while maintaining its core credit card processing with Stripe. Anthropic wants to unbundle parts of its payment stack. This will improve pricing logic, curb promotional fraud, and consolidate corporate banking operations.

This week’s reports

1️⃣Paypal 2Q’26 Results

2️⃣Amazon Bedrock AgentCore payments is now generally available

3️⃣Crypto Exchanges are Building Financial Super Apps for the Next Billion Users

4️⃣Strategic Roadmap for Agentic AI Adoption in Banking

5️⃣Finance & treasury without borders

6️⃣How do Europeans really bank, pay & invest in 2026?

7️⃣How instant payments are transforming the financial landscape

Paypal 2Q’26 Results

PayPal reported second-quarter adjusted earnings of $1.38 a share, beating analyst expectations of $1.28, and raised its full-year profit guidance as CEO Enrique Lores’s turnaround effort showed early signs of progress.

Revenue for the quarter rose 5% year over year to $8.68 billion. Analysts had expected $8.47 billion, according to The Wall Street Journal.

Revenue and Profit Metrics

Net revenues: $8.68 billion, up 5% reported or 3% on a currency-neutral basis.

Transaction margin dollars: $3.9 billion, a 1% increase (+3% excluding interest on customer balances).

GAAP net income: $1.10 billion (or $1.25 per diluted share), down from $1.26 billion ($1.29 per share) a year prior.

Read more

Fast-growing software platforms eventually hit a financial scale where off-the-shelf payment software creates operational friction. Anthropic saw its annualized revenue run-rate jump from $787 million to nearly $45 billion within five months by May 2026. The company reached a $965 billion valuation during its Series H funding round. When transaction volumes grow this quickly, small pricing inefficiencies and billing delays multiply into major operational expenses. Anthropic recently posted job listings for billing software engineers, fraud specialists, and treasury managers to evaluate developing proprietary financial tools.

Industry commentary quickly framed these job listings as a direct threat to Stripe. The full operational context presents a more nuanced reality. Anthropic is evaluating which specific financial primitives to build internally, while maintaining its core credit card processing with Stripe. Anthropic wants to unbundle parts of its payment stack. This will improve pricing logic, curb promotional fraud, and consolidate corporate banking operations.

Build vs. Buy in AI monetization

At early revenue stages, software startups happily pay a percentage fee to payment vendors in exchange for quick setup and low engineering overhead. Once annual revenues enter the tens of billions, paying standard platform fees alongside card processing costs becomes a significant margin drain.

Generative AI monetization differs sharply from traditional software subscriptions. Standard software charges a fixed monthly fee per seat. AI products charge based on dynamic compute metrics. These include token throughput, context window size, model size, and execution time. When customer billing relies on variable GPU compute usage, real-time metering precision directly impacts gross margins. Delayed billing synchronization or unmetered API calls quickly turn into lost revenue.

Anthropic’s open roles for billing platform software engineers highlight building custom foundational modules around third-party provider platforms. Anthropic is constructing proprietary entitlement engines, contract provisioning systems, and order management tools. These allow it to secure granular control over complex enterprise agreements and dynamic consumption tiers.

Read more

In this episode of WRAP UP, I sit down with Tamas Kadar, co-founder and CEO of SEON, to discuss how AI is making online fraud more sophisticated and what banks, fintechs and digital businesses can do to stop it.

We explore how deepfakes, synthetic identities and AI agents are helping fraudsters automate attacks at scale. Tamas explains why a face, document or video is no longer enough to verify a customer, and why fraud detection now needs to go beyond a single KYC check.

We discuss the signals that matter, including email, phone, IP, device intelligence, digital footprints and behavioural data. Tamas explains why risk teams need to monitor the full customer journey, from registration and onboarding to login and transaction, rather than rely on separate tools that only assess risk at one point in time.

We also cover account takeover through breached credentials and the growing threat of account handover, where real people open verified accounts that are later sold through Telegram and dark web marketplaces. These accounts can then be used for mule activity, scams and moving illicit funds.

Tamas shares how SEON has expanded its risk intelligence to more than 1,100 signals, how financial institutions can combine this data with their own internal AI models, and why dynamic friction is key: letting genuine customers move smoothly while introducing extra verification only when activity looks suspicious.

Finally, we discuss what banks and fintechs should do now to prepare for AI-driven fraud, including how to handle AI agents, why live proofs of concept matter when selecting a fraud platform, and why simple integration should be part of every vendor decision.

LEARN ABOUT SEON

LinkedIn: https://www.linkedin.com/in/sirojboboev/

Instagram: https://www.instagram.com/wrapuppodcast/

Tiktok: https://www.tiktok.com/@wrapuppodcast

X: https://x.com/sambobo

The reports highlight major shifts in financial services driven by AI, fraud prevention, agentic commerce, and blockchain-based payments. Agentic AI for Banking explores how autonomous AI can improve efficiency, personalization, and banking operations, while Banking on Banks 2026 examines how economic uncertainty, inflation, borrowing costs, and fraud are affecting European financial services. Agentic Commerce Pulse Report finds that agentic commerce is already gaining adoption, with security and data protection remaining key concerns. Meanwhile, Authorised Payment Fraud provides a policy framework for combating the growing fraud threat. The blockchain-focused reports argue for greater bank participation in public blockchain infrastructure and explore tokenized deposits and stablecoins as emerging forms of bank money.

Video of the Week

Deep Dive of the Week

Breaking Down PingPong’s Full-Stack Financial Infrastructure

When it comes to today’s cross-border landscape, legacy banking pipelines are largely fragmented. While existing payment orchestration provides connectivity, it does not solve the core structural problems of multi-entity reconciliation and settlement. Transaction routing systems introduce additional fees and latency, and each intermediary screens the payment on its own schedule, making settlement windows unexpected and unpredictable.

Add on another layer: data, and the problem becomes easier to miss and more complex to solve. In a multi-hop correspondent chain, structured remittance fields, ultimate debtor identifiers, and invoice references are stripped during format translations between intermediary core banking systems. That loss breaks straight-through processing for ERP systems and forces manual exception handling at the receiving end.

Replacing these intermediaries requires fully owned, vertically integrated rails. True scale in global payments comes from owning the whole stack rather than layering software over infrastructure someone else controls.

This week’s reports

1️⃣Agentic AI for Banking

2️⃣Banking on Banks 2026

3️⃣Agentic Commerce Pulse Report

4️⃣Authorised Payment Fraud

5️⃣Banks Should Validate Public Blockchains

6️⃣Tokenized Money
for Banks

7️⃣What Is The PayPal Mafia Building In 2026?

Agentic AI for Banking

Many financial services organizations have already implemented early AI technologies. Agentic AI represents the next step in the evolution of autonomous AI technology and a significant leap in its capabilities. But how do you bring it to your organization successfully?

The potential impacts of AI for financial services:

- Efficiency savings, innovation and $3 trillion in corporate productivity gains

- A 5.4% EBITDA improvement for the average company

- An average 2.3x return on agentic AI investments within 13 months

- Hyper-personalization that directly translates to stronger customer loyalty and reduced churn

In “Agentic AI for Banking”

This eBook is a guide for all financial services organizations looking to take agentic AI beyond pilots and proofs of concept into day-to-day operations. In it, you’ll discover:

What agentic AI is and how humans and AI agents work together

Read more

When it comes to today’s cross-border landscape, legacy banking pipelines are largely fragmented. While existing payment orchestration provides connectivity, it does not solve the core structural problems of multi-entity reconciliation and settlement. Transaction routing systems introduce additional fees and latency, and each intermediary screens the payment on its own schedule, making settlement windows unexpected and unpredictable

Add on another layer: data, and the problem becomes easier to miss and more complex to solve. In a multi-hop correspondent chain, structured remittance fields, ultimate debtor identifiers, and invoice references are stripped during format translations between intermediary core banking systems. That loss breaks straight-through processing for ERP systems and forces manual exception handling at the receiving end.

Replacing these intermediaries requires fully owned, vertically integrated rails. True scale in global payments comes from owning the whole stack rather than layering software over infrastructure someone else controls.

This is where PingPong Payments becomes an interesting case study. It holds 82 licences and authorisations worldwide, clears through direct connections to the payment schemes rather than third-party aggregation layers, and, in the markets where it holds its own regulatory permissions, issues local account details in its own name rather than a partner bank’s.

PingPong was originally built around the lifecycle of money rather than around separate payment products. Its infrastructure now processes more than $300 billion in annualised volume for over 750,000 active business clients across more than 200 markets. This puts it alongside some of the largest players in this space, including Airwallex and Checkout.com.

It is worth being clear about what that model costs, and the strength of the moat it creates. Licences take years to obtain and continuous investment to maintain; capital must be held against them, and direct scheme membership carries operational obligations an orchestration layer never has to think about. The argument for vertical integration is not that it is cheaper to build. It is that it is cheaper to run at volume, and that control over the payment becomes a product capability rather than a dependency.

The first hard problem starts before the payment

In older payment setups, identity and risk checks often sit before the transaction flow. A business submits its information, compliance systems run their checks, and anything uncertain ends up in a review queue. I have seen how quickly this becomes a product problem. A payment may be perfectly normal, but one weak signal can hold it up while someone investigates.

A full-stack payment engine handles this differently. Risk checks become part of the transaction flow. As a payment comes in, the system can pull company data from official registries and work through ownership structures to identify the people behind the business. Sanctions screening can happen at the same time. Much of this can run automatically within seconds.

That information can then influence what the payment engine does next. A company with a long history of verified activity may receive higher transaction limits and move through with little friction. A new business sending an unusual payment from a new device may face another check. The system can look at the connection source and compare the transaction with previous behaviour before deciding what action to take.

This is where I think the design of the payment stack matters. Compliance should be able to react to the risk of a specific payment instead of treating every transaction the same way. If something looks unusual, the engine can request a document or apply another check to that case in that moment, versus waiting until the end to decide. Meanwhile, verified payments can continue processing without being pushed into the same review queue.

What that buys is easy to state and hard to build. Fewer legitimate payments are stopped, so approval rates rise. Verification runs against live registry data rather than a document queue, so the gap between onboarding a client and clearing their first payment narrows. And the cost of compliance stops climbing in a straight line with volume, because the expensive checks are the ones the risk actually warrants. Screening becomes a variable cost, priced to the payment in front of it, rather than a fixed toll charged on everything that passes.

Virtual accounts and native clearing

Getting paid across borders is often harder than sending the payment.

A company may invoice a customer in another country and give them bank details connected to a pooled collection account. The customer sends an international wire. Two problems arrive with it. Correspondent banks may deduct fees along the way, so the sum that lands is smaller than the invoice and someone has to chase or write off the difference. Separately, remittance data is stripped in transit, so the receiving company still needs to work out which customer and invoice the payment belongs to. At scale the two compound. A short payment that cannot be matched to an invoice is a reconciliation problem and a collections problem in the same transaction, and finance teams end up solving both by hand.

This is why local collection infrastructure matters.

A full-stack provider like PingPong can issue virtual IBANs and local bank details linked to an individual business. A company in the US or Asia, for example, can collect money through local accounts in the UK or Europe without setting up a legal entity in every market. The payer gets familiar domestic payment instructions and can often pay through the local banking system instead of sending an international wire.

The difference sounds small. Operationally, it is significant.

When each virtual account is mapped to a customer or invoice, an incoming payment already carries useful information about where it belongs. The ledger can match the payment against an open invoice and pass that information into the company’s ERP. Treasury teams spend less time investigating incoming transfers and asking customers to provide payment references.

The same collection layer can support cards, bank transfers and more than 160 local alternative payment methods through one integration. Depending on the market, that can mean rails such as ACH in the US, SEPA in Europe, or Faster Payments and CHAPS in the UK. The payment method can change while the merchant keeps a common ledger and reconciliation layer.

Allocation is the part most providers leave to the platform. A multi-billion-dollar global marketplace launching across six European markets found the difficult problem was not moving money but dividing it. Every marketplace transaction contained at least three claims against the same incoming payment: a logistics fee, a platform fee, and a seller payout. Splitting and sweeping those automatically at the point of collection, rather than reconciling them afterwards, removes a standing workload from finance and operations teams. A single EMI licence in Luxembourg, for example, passported across 30 EEA states, covers every launch market without a separate banking relationship in each one.

There is a harder infrastructure question underneath all of this: who provides those local account details?

Many payment companies rely on banks or other licensed providers to supply accounts in markets where they do not have their own regulatory access. That can work well, but it creates another dependency in the payment chain. If a banking partner changes its risk policy or decides to stop serving a particular customer segment, the payment company has limited control over the outcome.

Building more of that infrastructure directly is expensive. It means obtaining the right licences, connecting to banks and payment systems, protecting customer funds and meeting capital requirements. The company must also maintain those controls as regulations change. But this investment gives the payment provider more control over how money is collected, held, and reconciled across markets.

The difference becomes more important once the money arrives

The ledger is where a cross-border payment platform starts to look like a treasury system.

At its inception, PingPong was built to help Asian e-commerce sellers collect marketplace revenue, convert currencies, and pay suppliers. As the company moved into B2B trade and then enterprise payments, those flows became more complex. Today, its broader stack includes multi-currency collection accounts, wallets and FX alongside payouts, issuing, and acquiring, all available as managed accounts through its API-based platform.

Think about a global marketplace. It might collect euros from European customers this morning, receive dollars later in the day, and have suppliers that need to be paid in several other currencies. Converting everything back into one base currency after every transaction creates unnecessary FX costs. A multi-currency ledger allows the business to hold those balances and use them again when the next payment needs to go out.

That changes the economics.

If a business already holds euros, a euro payment can be funded from that balance. The same principle can apply when a commercial card is connected to the wallet. Instead of converting the company’s base currency every time the card is used abroad, the platform can draw from the balance in the currency needed for that transaction.

Best Buy Canada is a useful example of what this unlocks. While running a third-party marketplace since 2016, its seller base was effectively domestic. In March 2025, it appointed PingPong as its first cross-border payment service provider of record, embedding payouts to international sellers through a single API integration. Best Buy Canada made the switch to lift gross merchandise volume, cut operational overheads, and improve the experience on both sides of the marketplace, and its marketplace leadership described the onboarding as fast. Wish and Wayfair run on this same infrastructure.

It’s worth noting that if a marketplace cannot pay a seller in that seller’s own currency, on rails that seller already uses, it cannot recruit that seller at all. The constraint being removed there is worth naming precisely, because it is not payment cost. It is seller supply. A marketplace that cannot pay a seller in that seller’s own currency, on rails that seller already uses, cannot recruit that seller at all. Cross-border payout capability is therefore a seller acquisition decision before it is a treasury one.

Liquidity management and embedded FX

But holding currencies creates another problem: liquidity.

PingPong has built a network that allows it to manage liquidity across markets. Once funds enter its network, the company can pay them out elsewhere, which has allowed it to support real-time, 24/7 payments for years. The infrastructure behind that includes 82 licences and around 200 banking partners, including relationships with Tier 1 banks.

FX gets more interesting once you move beyond a simple currency conversion.

Travel is a good example. An online travel company may find that the same airline ticket is priced differently across markets. Buying in another market can create an opportunity, but the company may have to settle with that airline at a later date. The ticket price may seem more attractive today, but the exchange rate may move against the company in that window before settlement.

So, the payment workflow needs to deal with that FX exposure at the moment the transaction is created.

PingPong’s approach embeds FX pricing into the company’s workflow and locks the relevant rate for the settlement period. This allows the purchase to be made with the FX cost known upfront, rather than leaving its treasury team to hedge the exposure separately.

Sabre faces exactly this on international bookings through its global distribution system, where the agency’s price is fixed at the point of booking, and the obligation to the supplier falls due days or weeks later. Pricing for the locked rate is aggregated live from more than ten liquidity partners and held through settlement.

I think this example explains why enterprise FX is easy to underestimate. A competitive spot rate matters, but enterprises often need much more than a cheap conversion. They need the FX product to understand the payment that sits behind it.

This is where the ledger, liquidity network, and FX engine start working as one system. The platform knows which currencies the business holds, which payments are coming in, and what needs to leave. In using smart-pricing engines, it’s much easier to decide when conversion is required and how best the FX exposure should be handled.

For an enterprise treasury team, this can remove a surprising amount of work from the payment process.

The final challenge is choosing how to deliver it

PingPong connects directly to schemes such as SEPA, Faster Payments, Bacs and CHAPS. It also uses SWIFT and local payment rails across different markets. That gives its routing engine several ways to move the same payment.

The practical effect is simpler than it sounds. A payment entering the network at one point can be paid out at another, without an international wire in between. Global liquidity management is what makes that possible.

I think the routing part is especially important. Cross-border payments are often discussed as if there is one best route. There isn’t. An enterprise may want the cheapest route for one payment and faster settlement for another. PingPong is using AI to make these routing decisions based on what matters to each customer, including the trade-off between speed and cost.

Risk controls must follow the payment through those routes. The useful question is not how much transaction monitoring a platform runs, but how few false positives it produces. Every held payment is a customer service cost, and at enterprise volume those costs compound faster than the fraud losses they prevent. A legitimate payment should keep moving. Transactions that need more information can be held for further checks.

Payout mesh and BPSP architecture

To execute mass international disbursements, vertically integrated engines run payout networks reaching more than 200 markets, paying out in 25 or more currencies over domestic clearing channels rather than correspondent chains.

That same infrastructure can also extend into enterprise supplier payments, where a different constraint emerges: many suppliers do not accept commercial cards. Interchange costs, merchant onboarding requirements and long-settled billing habits keep a large share of corporate spend outside the card rails altogether.

The Business Payment Solution Provider (BPSP) framework operated by the card networks exists to close that gap. Under it, an enterprise settles an invoice on its existing commercial card and the supplier receives an ordinary bank transfer, without ever onboarding as a card-accepting merchant. PingPong built its own version of this, Card to Account, with Visa, launching in May 2026 as one of three foundational providers selected for the programme. It is live in the UK, the EU and Hong Kong, with the US and Singapore following across 2026.

In practice the flow has three steps:

• The buyer settles supplier invoices on an existing commercial card programme, deferring cash outflow by up to 45 days without taking on new debt.

• The payment engine acquires the commercial card charge and routes an equivalent domestic bank transfer directly to the supplier’s domestic bank account via local clearing rails, reaching suppliers in more than 170 countries and 25 currencies, with funds landing between the same day and two days later.

• The supplier receives a standard domestic bank transfer without maintaining merchant card acquiring facilities or absorbing interchange discount rates.

Consolidating card acquiring, foreign exchange conversion and domestic clearing inside one platform removes the external gateway mark-ups that legacy BPSP models carry. It also shortens the chain itself. The fewer external systems that sit between the card being charged and the money reaching the supplier, the fewer places a payment can stall and the less reconciliation work lands on the treasury team at the end of it.

The infrastructure advantage

Software is becoming the easier part of cross-border payments.

A well-funded team can build APIs and connect payment providers relatively quickly. Nothing in the software layer is genuinely defensible. Given enough time and money, any payment interface can be replicated, and product development is getting even faster with AI.

The harder part takes years.

PingPong participates directly in schemes such as SEPA, Faster Payments, Bacs and CHAPS. That infrastructure gives it more control over how payments move and how liquidity is managed across markets.

The banking relationships are particularly difficult to copy. They are built up over years and granted on trust, demonstrated capability, and confidence in the strength of a counterparty’s compliance and risk controls. A new entrant with unlimited capital can buy engineers. It cannot buy a track record.

This is where I think the gap between a payment interface and a full-stack provider becomes clear. The API can be built quickly. Licences cannot. Neither can years of compliance history or the banking relationships needed to support global collections and payouts.

For an enterprise choosing cross-border infrastructure, those layers matter because they determine what the provider can actually control when money starts moving.

Disclaimer:

Fintech Wrap Up aggregates publicly available information for informational purposes only. Portions of the content may be reproduced verbatim from the original source, and full credit is provided with a “Source: [Name]” attribution. All copyrights and trademarks remain the property of their respective owners. Fintech Wrap Up does not guarantee the accuracy, completeness, or reliability of the aggregated content; these are the responsibility of the original source providers. Links to the original sources may not always be included. We use AI tools to brainstorm, draft, edit, and summarize, but the ideas, analysis, and connections Sam draws are his. Every piece is verified and written in Sam’s voice. For questions or concerns, please contact us at sam.boboev@fintechwrapup.com.

Tokenized equities could change that.

In this episode of WRAP UP, I sit down with Mark Greenberg, Chief Commercial Officer of Kra…

Read more