logo

Reflections on China's AI journey

Reflections on China's AI journey

China’s AI and digital finance evolution reflects decades of institutional learning, policy alignment and platform maturity. From Alipay’s trust revolution to AI-driven risk systems, the 2025 China AI Innovation Study Tour showcased how banks, techfins and regulators now innovate within an integrated ecoystem.

China’s path to artificial intelligence (AI) and digital finance has been built over more than two decades, shaped by platforms, regulation and financial institutions adapting in step. What the multinational delegation of the 2025 China AI Innovation Study Tour observed was not a sudden leap forward, but the cumulative result of historical milestones, regulatory corrections and institutional learning.

The journey began in the late 1990s and early 2000s with the rise of internet players such as Baidu, Alibaba and Tencent (BAT). Their early platforms created the digital commerce ecosystem, but the real breakthrough was in payments. Alipay, launched in 2004, solved the problem of trust in online transactions and rapidly became the de facto method for digital commerce. This was followed by WeChat Pay in 2013, which embedded peer-to-peer transfers into daily life and accelerated mobile payments to mass adoption.

The 2010s marked the era of the techfins, where technology companies expanded into financial services, often directly competing with banks. The launch of Yu’e Bao in 2013, Ant Group’s money market fund that swept up idle balances from Alipay, illustrated the disintermediation threat — growing into the world’s largest fund within a few years. Banks suddenly faced competition not just from each other but from platforms with hundreds of millions of daily users.

At the same time, super-apps and mini-programmes matured, with finance embedded alongside lifestyle, shopping, ride-hailing and entertainment. Banks could not ignore this reach. By the mid-2010s, many began distributing products through mini-programs inside Alipay and WeChat, and more recently through video platforms such as Douyin, which has evolved into a powerful channel for commerce and financial product discovery.

By the late 2010s, regulation stepped in to restore balance. Authorities tightened rules to enforce financial stability, consumer protection and data privacy, ensuring that TechFins could not grow without accountability. This recalibration marked the beginning of a shift from competition to collaboration. Today, platforms and their technology arms — such as Ping An’s OneConnect, Ant Group’s technology solutions and Tencent Cloud — work as infrastructure providers to banks, especially smaller institutions that cannot invest in building advanced systems independently.

This transition is guided not only by market forces but also by national policy priorities. As Li Lin of Shanghai Pudong Development Bank (SPDB) explained during the tour, China’s financial system is being restructured around the “five major essays”: technology and digitalisation, green finance, inclusive finance, pension finance and cross-border finance. These essays reflect the state’s emphasis on aligning financial development with broader economic and social goals. For banks, this means embedding digital and intelligent finance at the core of operations, while also ensuring that innovation supports inclusivity, sustainability, ageing demographics and China’s global connectivity.

OneConnect echoed this perspective, showing how Ping An’s technology arm develops AI-driven tools for digital know-your-customer (KYC), portfolio modelling and compliance not only for Ping An itself but also for partner banks in China and overseas. This illustrates how the policy essays are operationalised: technology subsidiaries of major groups act as enablers for smaller institutions, allowing them to adopt intelligent finance solutions more quickly and in line with regulatory objectives.

Delegates saw how this plays out in practice: OneConnect exporting AI and risk management solutions across Asia and Africa. MYbank enabling small and micro enterprises to access loans in three minutes, with zero human intervention. WeBank scaling blockchain-enabled data sharing and a cloud-native stack to both its 420 million customers and third-party institutions.

The relationship has matured into one of co-dependence: platforms provide distribution and digital rails, while banks bring balance sheets, compliance and risk expertise.

The newest stage of the journey is defined by the systematic embedding of AI. AI now powers not only chatbots and customer engagement, but also credit underwriting, fraud detection, anomaly monitoring, code generation, compliance surveillance and portfolio simulation. Delegates witnessed MYbank’s “bird systems,” Lexin’s LexinGPT anomaly detection and simulation framework, Douyin’s Doubao large language model, and Ant Group’s privacy-preserving AML engines. These examples showed how AI frameworks are being institutionalised into processes, rather than remaining pilot projects.

Participant feedback underscored these points. Many noted that China’s model integrates infrastructure, platforms and regulation into a coherent ecosystem where innovation is entrepreneur-driven but nationally aligned. Several remarked on the maturity of compliance systems, especially Ant Group’s regtech capabilities, and the technical resilience demonstrated by indigenous solutions such as

WeBank’s distributed database and Huawei’s hardware. Others reflected on the cultural drivers — from Douyin’s young and energetic workforce, MYbank’s average employee age of 31 combining digital fluency with risk maturity, to Huawei’s ethos of continuous improvement — that sustain scale and innovation.

The consensus was that China’s AI innovation journey is unique: platforms and banks are no longer in zero-sum competition but in symbiosis, with regulation as a balancing force and indigenous technology as a strategic anchor. It is a model difficult to replicate outside China, but one that offers insights into how AI, data and platforms can reshape finance at scale.