Ant International Launches FalconTST 2.0 as Global Banks Adopt AI for FX Management

August 21, 2026 • FinTech News

Ant International Launches FalconTST 2.0 as Global Banks Adopt AI for FX Risk Management

Ant International has launched Falcon Time-Series Transformer (TST) Model 2.0, an artificial intelligence model designed to improve foreign-exchange forecasting and liquidity management for financial institutions.

The new model is already being used by major global banks, including Barclays, Citi, Deutsche Bank and Standard Chartered. Reuters also reported that Ant International has partnered with six large banks, including Citi, HSBC, Deutsche Bank, Standard Chartered and Barclays.

The launch highlights a growing shift in banking. Financial institutions are moving beyond general-purpose AI and adopting specialised models for specific financial tasks.

FalconTST 2.0 Targets FX and Liquidity Management

FalconTST 2.0 focuses on time-series forecasting. This makes it suitable for financial data that changes continuously.

Ant International designed the model to analyze data such as transaction flows, cash balances, and currency exposure. These forecasts can help businesses estimate when they will need liquidity and in which currencies.

The technology initially supported Ant International’s own cash flow and FX management. The company now works with banking partners to integrate FalconTST into their financial systems.

Kelvin Li, General Manager of Platform Tech at Ant International, said specialized financial AI can offer advantages over general-purpose models. Reuters reported that Ant International believes more accurate forecasts could cut FX hedging and allocation costs by more than 60%.

FalconTST 2.0 Reports More Than 93% Forecast Accuracy

Ant International says FalconTST 2.0 achieved a Mean Absolute Scaled Error (MASE) score of 0.666 on a global public benchmark for time-series foundation models.

The company says this result placed the model at the top of the benchmark leaderboard.

Ant International also reports that FalconTST 2.0 has achieved more than 93% forecasting accuracy consistently in its current financial applications.

These figures are company-reported results. They should therefore be viewed in the context of Ant International’s own testing and deployments.

Barclays, Citi and Standard Chartered Integrate the Model

The latest version is already connected to several banking platforms.

Barclays has integrated FalconTST into its BARX NetFX platform. Citi combines the model with its Fixed FX Rates solution.

Standard Chartered uses FalconTST alongside its SCALE FX system. The collaboration forms part of the bank’s participation in the Monetary Authority of Singapore’s PathFin.ai programme.

Standard Chartered previously said its integration with FalconTST could forecast Ant International’s FX exposure with more than 90% accuracy. The bank said the technology could help improve FX risk management and reduce hedging costs.

Why Specialized AI Matters for Banks

Most generative AI systems focus on language, images or other unstructured information.

Financial institutions also deal with large volumes of numerical data. Currency positions, transaction values, liquidity needs and settlement flows can change throughout the day.

That creates a different AI challenge.

FalconTST is built specifically for time-series data. Ant International says the model can work across different time frequencies, from second-level payment data to hourly treasury requirements and daily or monthly business indicators.

The model also aims to distinguish missing data from genuine zero values. This matters in financial forecasting because a lack of transactions does not always mean that demand has fallen to zero.

AI Could Help Reduce Cross-Border FX Costs

Foreign-exchange risk remains a major issue for companies operating internationally.

A business may receive revenue in one currency while paying suppliers in another. Currency movements can then change the value of expected cash flows.

Better forecasting can help treasury teams prepare for those changes.

Ant International says FalconTST 2.0 can help businesses decide how much liquidity to hold, when funds will be needed and which currencies they should hold.

For banks and payment companies, those decisions can affect hedging, liquidity and capital efficiency.

From Payments to Broader Business Forecasting

Ant International is also looking beyond foreign exchange.

The company says FalconTST 2.0 could support e-commerce supply-chain demand forecasting and aviation operations management.

That could expand the model beyond banking and payments into other industries that rely heavily on changing numerical data.

The strategy reflects a broader trend in financial technology. AI is increasingly moving into the operational systems that manage money rather than remaining limited to customer-facing applications.

Ant International Expands Its AI Strategy

The FalconTST launch comes as Ant International continues to expand its global technology business.

Reuters reported that the company raised $1.2 billion in its latest equity fundraising as it looks to support international expansion.

The company is positioning specialized AI as part of its wider financial infrastructure.

For banks, the appeal is practical. More accurate forecasts can help them manage liquidity and currency exposure. For cross-border payment providers, the technology could also help reduce the financial friction created by volatile exchange rates.

What the FalconTST 2.0 Launch Means

Ant International’s latest AI model shows how financial AI is becoming more specialised.

Banks do not simply need AI that can generate text. They need systems that can analyse financial data and support real-time decisions.

FalconTST 2.0 targets that requirement directly.

With major banks already integrating the technology, Ant International is positioning its time-series AI as a tool for FX forecasting, liquidity management, and financial risk control.

The next stage will be whether these systems can deliver consistent savings and operational improvements at a larger scale.

For now, the adoption of FalconTST 2.0 by major banking institutions marks another step toward AI becoming part of the core infrastructure behind global payments and treasury management.

Citi, HSBC, StanChart adopt Ant International’s forex AI tool