On September 18, Lu Lei, Deputy Governor of the People's Bank of China, delivered a keynote address at the China-ASEAN Financial Cooperation and Development Forum, focusing on the theme of "Deepening AI Cooperation in the Financial Sector Between China and ASEAN."
Reflecting on the trajectory of technological and economic history, Lu noted that every major technological revolution has triggered profound socio-economic shifts while leaving an indelible mark on the evolution of the financial industry. He observed that artificial intelligence is now fundamentally reshaping the operational logic of financial institutions and the mechanics of financial markets, with its "double-edged sword" effects becoming increasingly apparent. This reality, he argued, calls for collaborative efforts and shared capabilities across borders to steer AI development along a responsible and constructive path.
Drawing on China's practical experience, Lu outlined how the People's Bank of China has employed strategic planning, pilot programs, and standardized frameworks to guide financial institutions in adopting AI responsibly. This approach has cultivated a distinctive application model characterized by a progression from peripheral to core functions, from assistive to collaborative roles, and from internal to external deployment. Concurrently, the development of next-generation digital yuan infrastructure has laid the monetary and systemic groundwork for integrating AI and smart contract technologies. These initiatives, Lu emphasized, have proven instrumental in mitigating risks, strengthening regulatory oversight, and fostering high-quality development, while also offering valuable insights for knowledge-sharing between China and ASEAN nations.
Lu highlighted that emerging AI technologies, particularly large language models and intelligent agents, serve as powerful enablers for financial services yet introduce novel risks such as algorithmic opacity and model hallucination. He warned that while AI can discern the underlying logic of human behavior, humans often struggle to penetrate AI's internal decision-making processes. This "asymmetric understanding" between humans and machines, he cautioned, could pose deep-seated challenges to the financial sector and underscores the critical importance of robust safety governance.
In the context of China's practice, Lu reiterated the commitment to balancing development with security, guiding financial institutions to adopt AI in a measured and orderly manner. This involves harnessing AI's full potential in processing vast datasets, recognizing complex patterns, and generating multimodal content, while simultaneously leveraging human strengths in value judgment, ethical decision-making, and creative innovation. The goal, he stated, is to ensure that AI applications in finance remain safe, reliable, and controllable.
Expressing a willingness to collaborate with ASEAN countries, Lu proposed joint efforts to assess novel financial risks arising from AI technologies, exchange practical experiences in model lifecycle management and algorithmic risk prevention, and build safety protection systems aligned with digital and intelligent development. Such collaboration, he affirmed, would fortify the "AI plus finance" security framework.
Lu underscored that application serves as the true test of technological innovation, with real-world scenarios providing the fertile ground for deployment. From China's perspective, the focus has consistently been on application-driven and scenario-led approaches, starting from specific business needs and addressing practical pain points. By coordinating core elements like business operations, technology, data, and talent, China has advanced AI adoption in a context-specific, incremental, and iterative manner, yielding a portfolio of replicable and scalable AI finance cases that have enhanced operational efficiency and service quality.
Looking ahead, Lu expressed eagerness to deepen collaboration with ASEAN on scenario co-construction and outcome sharing, broadening the scope and depth of technology application, fostering an integrated regional innovation ecosystem, and identifying new growth drivers for financial cooperation. The ultimate objective, he noted, is to translate AI's technological dividends into tangible development benefits.
Addressing resource constraints, Lu acknowledged that with large model parameters escalating under the influence of scaling laws, resources such as computing power, data, and talent remain scarce, with significant disparities across institutions. Relying solely on individual efforts, he cautioned, risks widening the intelligence gap between larger and smaller entities while exacerbating third-party dependencies and concentration risks.
Drawing on China's approach, Lu highlighted the commitment to intensive efficiency and co-construction with shared benefits. This includes advancing a national pilot base for AI financial applications, launching "data element X" trials in the financial sector, and building industry platforms for co-creation, supply-demand connectivity, model training, data sharing, and standard transformation. These efforts aim to optimize resource utilization across the entire financial industry.
Lu voiced readiness to deepen cooperation with ASEAN in co-building AI financial application infrastructure, exploring shared multilingual high-quality financial datasets, secure computing power sharing mechanisms, and multimodal financial domain model matrices. Such partnerships, he said, would cultivate an open, inclusive, and mutually beneficial ecosystem, advancing sustainable development that balances quality, efficiency, and scale.
If data, computing power, and models form the foundational pillars of "AI plus finance," then safe, efficient, and interconnected digital financial infrastructure serves as the channel connecting technology to real-world impact. Lu noted that the People's Bank of China is currently enhancing the digitalization and intelligence of financial infrastructure in line with digital central banking requirements and high-standard financial opening.
As an illustration, Lu cited the newly launched "Digital Yuan New Cross-Border Trade Settlement and Investment Financing System, known as the Shuda Da Li Fu Service." This system leverages the digital yuan cross-border payment platform, or CBETS, as its settlement and investment base, integrating with regional digital trade platforms that act as enterprise-facing service windows. It streamlines cross-border trade, shipping logistics, and investment financing by unifying ledgers to achieve a "three-flow integration" of information, data, and documentation streams.
The system also incorporates cutting-edge technologies like intelligent agents and digital yuan smart contracts, with the latter automating contract execution based on agreed terms to program financing and payment logic, reducing compliance costs and transactional friction. Intelligent agents, meanwhile, assist in manual review processes, verifying trade document authenticity and enabling proactive risk alerts. Lu expressed confidence that this framework would deliver an efficient, secure, inclusive, and closed-loop digital yuan trade settlement and financing channel for regional economic exchanges.
Lu concluded with a philosophical reflection, stating that while AI technology itself is neutral, the manner and direction of its application are decisive in determining success or failure in financial use cases. In China's experience, the guiding principle has been to prioritize the enhancement of people's well-being, using AI to broaden financial service channels and reduce costs while unlocking its potential in extending outreach, lowering financing expenses, optimizing livelihood services, and empowering rural revitalization.
Citing examples, Lu mentioned banks that utilize large model technologies to refine credit assessments for micro, small, and agricultural entities, addressing financing challenges posed by a lack of traditional collateral and elevating financial services in county and rural areas. Other banks employ intelligent interaction tools to deliver age-friendly and accessible financial services to the elderly and residents in remote regions.
Lu expressed a strong desire to share China's best practices in leveraging AI for public good with ASEAN nations, utilizing technology to bridge regional digital and intelligent divides and ensuring that groundbreaking financial innovations benefit people and market entities across all countries.
Comments