TOKYO, January 26, 2026 - Preferred Networks, Inc. (PFN) has been selected as the digital partner for Japan International Cooperation Agency (JICA)'s proof-of-concept (PoC) project titled Proof of Concept of an AI-based liquidity forecasting solution with National Bank of Cambodia, in which PFN is developing an AI model that predicts liquidity in Cambodia's local currency (riel) for the National Bank of Cambodia (NBC), beginning September 2025. This PoC is part of JICA's DXLab initiative, which aims to support emerging economies by collaborating with Japanese and local companies with strengths in digital technologies. The project is scheduled to continue through February 2026.

National Bank of Cambodia

Liquidity refers to the ease with which an asset can be converted into cash, and accurately predicting its overall trends serves as a critical basis for central banks' monetary policies such as interest rate adjustments, market interventions and government bonds. In this PoC, PFN's team specialized in quantitative finance will train a proprietary AI model with historic data related to riel liquidity so that it can predict liquidity over specific future periods. Accurate forecasting of riel liquidity will enable NBC to establish a data-driven foundation for monetary policy, thereby contributing to the country's goal of achieving independent monetary policy.

Due to the high degree of dollarization in Cambodia, NBC is currently facing challenges in managing the Cambodian riel supply and implementing effective interest rates as the bank is unable to issue U.S. dollars. To address this, NBC has been actively promoting riel usage. NBC has proactively applied cutting-edge technologies to address the issue, including the blockchain-based Bakong payment systems to modernize monetary policy and payment infrastructure.

PFN maintains a dedicated team for quantitative finance that researches and develops AI-based solutions specifically for the financial sector, with multiple papers published at specialized conferences. The team has extensive research experience in financial time series analysis, which is applied in this PoC, and recently developed a large language model specialized for use in finance.

Upon successful completion of this PoC, NBC will look into full-scale implementation of the AI model and aim to set an example for good practice of AI utilization at a central bank, while also exploring potential deployment in Southeast Asia and beyond.