August 2026 Summaries

Airwallex describes how rapid growth and the complexity of global payments created fragmented financial data, with millions of micro-transactions, differing product schemas, and inconsistent definitions across markets. Following a migration from BigQuery to Databricks, the company made Finance responsible for defining decision-driven use cases and built its reporting model around the reconciled Airwallex ledger. Finance teams across tax, treasury, controllership, and FP&A collaborated on a shared data dictionary, standardized nomenclature and table structures, and expanded ERP account tagging to create a governed, business-wide source of truth. This foundation provides AI tools with the context needed to interpret financial fields correctly, enabling faster tasks such as attributing regional operating-cost changes through agents used in platforms including Cursor, Claude, and Hex. The company argues that the main obstacle to using AI in finance is often data readiness rather than data sensitivity, while noting that maintaining a unified data model remains an ongoing challenge as new business data emerges.

Airwallex’s global accounts, foreign-exchange capabilities, regulatory controls, and local payment networks already allow businesses to collect, convert, and distribute funds efficiently in more than 120 countries, reducing the practical need for stablecoins in many conventional cross-border payments. Although stablecoins can help onchain businesses and users in unstable banking or inflationary environments, they still require regulated infrastructure to convert digital dollars into local currencies and complete supplier, employee, or contractor payouts. The company argues that AI agents may create a stronger use case, as stablecoin wallets can impose fixed spending limits and support low-value, high-volume micropayments that are inefficient under card-fee models. With evolving U.S. and European regulations and growing deployment of AI agents, Airwallex is integrating stablecoin support where permitted, investing in settlement provider Metal, and aiming to let customers choose between stablecoins and local currency based on payment needs, particularly for agent transactions and slower banking corridors.

AI agents reportedly drove substantial online retail activity during Black Friday 2025, aided by established consumer protections such as returns, chargebacks, and card-network liability rules, but corporate finance remains more cautious because erroneous payments can cause financial losses, compliance breaches, or regulatory consequences. A Forrester Consulting study commissioned by Airwallex found that only 11% of core finance workflows operate without human involvement, despite strong interest in AI adoption and projected productivity benefits, because leaders also prioritize controls, audit readiness, and compliance. The central challenge is explainability: unlike deterministic spreadsheet calculations, AI agents may reach differing decisions in similar circumstances, requiring traceable records of their actions, permissions, rationale, and accountability. The article frames this as “Know Your Agent,” extending Know Your Customer principles to autonomous software, while arguing that safeguards must cover data, payments, treasury, reconciliation, fraud monitoring, and transaction controls rather than the AI model alone. It notes that fragmented data, manual workflow steps, and limited AI-process experience constrain adoption, and suggests that many organizations expect vendors to provide certified platforms and decision records while finance teams retain responsibility for configuration and oversight.

Rapid changes in tariffs, court rulings, currency markets, and emerging payment technologies are expanding treasury’s role beyond traditional liquidity, cash, and foreign-exchange management into a strategic function responsible for supplier decisions, trade routes, pricing, market entry, and risk response. The rise in effective US tariff rates, followed by a Supreme Court ruling invalidating certain IEEPA tariffs and potentially triggering up to US$175 billion in refunds, illustrates how legal and policy developments can create urgent decisions requiring current data on cash, exposure, and corporate structures. Businesses affected indirectly through customers, supply chains, infrastructure costs, or extended payment terms also need treasury to assess the resulting implications for working capital and margins faster than monthly reporting cycles allow. The passage argues that connected financial infrastructure, including real-time multi-entity cash visibility, multicurrency accounts, integrated FX and payments, and current scenario modelling, can reduce delays and enable faster action, while noting that such systems cannot remove regulatory constraints such as export controls or capital restrictions. It also identifies autonomous AI agents initiating payments as an emerging governance challenge, as treasury may be accountable for permissions, spending limits, reconciliation, and liability for transactions it did not manually approve.

After discussions with more than 20 private-equity-backed CFOs, an Airwallex finance leader found that AI adoption in Finance has progressed slowly, constrained by lean teams, risk aversion in controllership, and poor data hygiene, while current experimentation is concentrated in FP&A tasks such as variance analysis and management-reporting presentations. The author argues that CFOs should consider redesigning Finance operations rather than merely adding AI to existing workflows, focusing first on the quality of the data foundation, a detailed understanding of teams’ work and manual processes, and an operating model that enables people to use new tools effectively. Airwallex plans to share lessons from its process mapping, team and operating-model evolution, and AI roadmap as it seeks to build a scalable Finance function.

Airwallex’s latest product updates focus on automating finance operations across payments, billing, spend, and embedded finance. Payment enhancements include configurable surcharge or discount rates with regional compliance controls, an AI-powered Fraud Attack Prevention Agent for detecting and blocking card-testing attacks, and Wero support to reach more than 55 million European account-to-account payment users. Billing additions include a QuickBooks Online integration for synchronizing invoice records, reusable branding themes for invoices and checkout, and embedded checkout elements that let customers manage billing workflows on a business’s own website. Spend updates introduce an expense policy agent that can automatically approve or escalate expense submissions, alongside APIs for synchronizing accounting reference data with ERPs. Platform customers with Netherlands-based entities can also make payouts directly to eligible Visa and Mastercard cards, while early-access features are available through Airwallex support or account representatives.

Airwallex now enables eligible merchants in 35 markets to offer Affirm to shoppers in the United States, providing a buy now, pay later option that lets customers pay in four interest-free installments or through monthly plans with costs disclosed upfront and no late or hidden fees. Affirm may be particularly useful for higher-value categories such as electronics, furniture, and travel, where flexible payment options can influence purchasing decisions. As of the quarter ended March 31, 2026, Affirm reported nearly 27 million users, 22% year-over-year active consumer growth, 44% active merchant growth, and US$11.6 billion in gross merchandise volume. Airwallex allows merchants to add Affirm without establishing a US entity or separately managing another provider, using a single integration across options including hosted checkout, APIs, mobile SDKs, Shopify, and payment links, alongside its wider portfolio of more than 160 global payment methods.

Startup Finance 30 is an annual, merit-based global recognition program launched in 2026 by Airwallex to highlight 30 finance leaders whose work has significantly supported growth at venture- or growth-backed private startups, generally from Series A through pre-IPO. Eligible nominees must lead or own the finance function, have an established track record, and meet integrity standards, while public, pre-revenue, and unresolved governance-issue companies are excluded. Candidates are independently scored on their impact on company growth, strategic responsibilities beyond finance, ability to build scalable teams and systems, use of AI and technology to improve finance operations, and influence on the broader finance ecosystem. An independent committee of finance executives from Notion, Vercel, Deel, and Airwallex reviews shortlisted candidates and ratifies the final 30, with safeguards including an editorial firewall, no participation fees or paid placements, customer-agnostic review, conflict recusals, and public-record verification of honorees.

AI is reducing the defensibility of software features and workflows, making embedded financial infrastructure a more durable source of differentiation for platforms. While payment acceptance remains important, the greater strategic challenge involves managing post-transaction money flows, including fee deductions, fund splitting, settlement, balances, currency conversion, reconciliation, and payouts across users and markets. Strong embedded finance can improve user trust, operational efficiency, international scalability, and platform margins by providing transparent, flexible control over funds and reducing reliance on disconnected providers. The passage presents Airwallex’s connected-account infrastructure as a solution that supports multi-currency accounts, global pay-ins and payouts, FX conversion, fund routing, and configurable integrations with existing payment providers, positioning financial capabilities as an extensible foundation for platforms as software development becomes easier to replicate.

E-invoicing is the exchange of structured digital invoice data, typically in formats such as UBL or CII, between business systems through approved networks or platforms, rather than PDFs designed only for human reading. Governments worldwide are increasingly mandating or developing e-invoicing and related reporting rules to improve tax compliance, reduce fraud, and provide timely transaction data, while businesses can benefit from automated validation, faster approvals and reconciliation, stronger security, and improved invoice traceability. Requirements vary substantially by country, transaction type, business size, format, network, and reporting model: Italy and many Latin American countries already have broad mandates, France will begin requiring receipt of e-invoices in September 2026, Germany is phasing in issuance requirements through 2028, the UK plans a mandate for 2029, and Singapore currently requires separate invoice-data reporting through its Peppol-based network. Businesses are advised to assess relevant jurisdictions and entities, confirm local rules and deadlines, improve master data quality, consider cross-border and non-B2B obligations, and ensure finance systems can generate, receive, validate, store, and process structured invoices. Airwallex plans to introduce support beginning in France, allowing eligible VAT-registered French businesses to receive and process e-invoices through its Spend and Billing products, with additional market coverage expected over time.

AI adoption in finance is constrained less by model capability than by fragmented data, workflows, and infrastructure that prevent systems from accessing the trusted context needed to execute decisions autonomously. A Forrester Consulting study commissioned by Airwallex in June 2026 found that only 11% of finance workflows operate without human intervention, while 65% of finance leaders identified inconsistent data and 61% cited cross-region and entity integration complexity as barriers to scaling AI. The piece argues that autonomous finance requires a unified view of balances, transactions, approvals, entities, and settlements, alongside governed workflows and operational infrastructure such as licences, payment rails, and security controls. It compares the shift to electrification, suggesting that organisations gain the most when they redesign underlying operations rather than merely add new technology. Airwallex presents its global licensing and payment-network infrastructure, as well as Minor Hotels’ reported US$7 million annual savings from consolidating payments, as examples of how integrated platforms can reduce complexity while creating the data foundation needed for secure AI-driven financial operations.