Ant International Rolls Out First‑Ever Full‑Stack AI‑Native Suite for Global Finance Operations
- Nishadil
- September 19, 2026
- 0 Comments
- 3 minutes read
- 6 Views
- Save
- Follow Topic
Ant International unveils industry‑first AI‑native stack covering payments, accounts, FX, treasury and growth for worldwide businesses
In its biggest product launch to date, Ant International introduced a full‑stack AI‑native platform that blends payment, account, FX, treasury and growth tools into a single, secure solution for global merchants.
Shanghai & Singapore – 18 September 2026 – Ant International surprised more than a thousand fintech leaders at its annual VOYAGE merchant event by pulling back the curtain on a truly ambitious upgrade: a full‑stack, AI‑native suite that handles everything from payments to foreign‑exchange forecasting, treasury management and growth‑driven analytics.
The heart of the new stack is a pair of home‑grown foundation models. First, the Antom 3‑in‑1 Transformer – a payment‑focused model that can read sequential behaviour, tabular records and graph‑based relationships all at once. With over 10 billion parameters and the ability to chew through roughly 90 trillion data points each year, it powers fraud detection, scam‑prevention and real‑time success‑rate optimisation in one go.
Second, FalconTST, a time‑series transformer built for FX and liquidity forecasting. By pulling data from finance, retail, energy and travel sectors, the model (now at 8.5 billion parameters) claims more than 93 % accuracy, trimming hedging costs for corporations by up to 60 % in early pilots.
Security isn’t an afterthought. Ant International layered a second “shield” on top of the models, deploying a dynamic Know‑Your‑Agent (KYA) framework via its Agentic Mobile Protocol (AMP) and a unique AgentSafePay guarantee that backs funds against agent‑specific risks.
What really makes the offering feel “agent‑ready” is the embedded autopilot functionality. Using natural‑language prompts, users can launch payment runs, reconcile accounts, or run treasury simulations without touching a single line of code. The suite even introduces the world’s first Account‑for‑Agent (AFA) product, letting businesses treat agents as first‑class financial entities.
Numbers from Ant International’s own merchant network speak loudly. Roughly 89.5 % of merchants using the Antom payment service have already added AI agents to their workflow, and 81.4 % of payment‑related tasks now receive AI assistance. Early adopters report charge‑back rates dropping as much as 87 % after switching to the new subscription‑based AI layer.
Beyond payments, Ant International’s ecosystem spans Alipay+, Antom, WorldFirst and Bettr, linking 150 million merchants to over 2 billion consumer accounts across 210 markets. Daily transaction volume tops 25 million, underscoring the scale at which the new AI stack will operate.
“The future fintech leader needs a trust‑first infrastructure and a set of foundational FinAI models that blend deep data with domain expertise,” said CEO Peng Yang during the launch. “Our full‑stack, AI‑native solutions give businesses the efficiency they need today and the agentic capabilities they’ll rely on tomorrow.”
With this sweeping upgrade, Ant International hopes to shift the industry from fragmented tools to a unified, AI‑driven workflow – a move that could redefine how global enterprises manage money, risk and growth in the coming years.
- India
- Pakistan
- Business
- News
- Australia
- BusinessNews
- Singapore
- China
- NewZealand
- Japan
- SriLanka
- SouthKorea
- Bhutan
- Malaysia
- Indonesia
- Maldives
- HongKong
- Afghanistan
- Nepal
- Bangladesh
- Thailand
- Mongolia
- Brunei
- Philippines
- Cambodia
- Fiji
- Pdt
- FintechAi
- TreasuryManagement
- AntInternational
- PaymentAutomation
- FxForecasting
- Antom3In1Transformer
- Falcontst
- AiNativeFinancialSolutions
- AgenticMobileProtocol
- GlobalMerchants
Editorial note: Nishadil may use AI assistance for news drafting and formatting. Readers can report issues from this page, and material corrections are reviewed under our editorial standards.