Ant rolls out Falcon open-source AI model to help users forecast cash flow, forex exposure



Ant International has released an open sourced foundational AI model called Falcon to help customers forecast cash flow and foreign-exchange exposure, as one of the world’s largest fintech companies puts artificial intelligence front and centre in its priorities to spur growth.

The Falcon TST model could help customers including airlines, online travel agents and e-commerce platforms forecast their cash flow and currency exposure hourly, daily and weekly with more than 90 per cent accuracy, executives said.

The model could handle “a big part of liquidity management and [forecasts’] multicurrency needs”, helping customers to “save cost by 60 per cent”, said Ant International’s CEO Yang Peng, during a panel discussion at the 2025 Singapore Fintech Festival.

The foundational model, available on GitHub and Hugging Face, is the most important AI project for the Singapore-based company, which handled nearly US$1.5 trillion in money flow in the past 12 months, 36 per cent more than in 2024. The model enables businesses to make various other time-series forecasts, such as weather patterns, calendar events, financial market fluctuations and cross-border traffic data.

The digital payment and fintech provider has also developed a payment risk control model called “Shield” 3-in-1 transformer model.

“It’s now protecting the 22 million transactions that we handle every day,” Yang said, adding that the pool of daily applications – comprising everything from QR codes to open banking and card processing – was “big enough” to let it detect fraud.

  • Related Posts

    How Changan Automobile’s design-driven brand strategy shapes its global expansion plans

    Long seen as one of China’s traditional “big four” state-owned vehicle manufacturers, Changan Automobile, which has a history that dates back to 1862 but began to produce its own eponymous-branded…

    Continue reading
    Alibaba teases new Qwen previews, highest-ranking Chinese AI models on Arena

    The new model previews, the Qwen3.7-Max-Preview and the Qwen3.7-Plus-Preview, ranked 13th globally in text capabilities and 16th in vision capabilities, respectively, benchmark firm LM Arena said on Tuesday. The versions…

    Continue reading

    Leave a Reply

    Your email address will not be published. Required fields are marked *