The key question is whether it can narrow the gap in financial literacy across different groups.
Photo from Jiemian News
by ZHANG Xiaoyun
AI is spreading quickly through every stage of wealth management, from boosting research efficiency to reinventing advisory models, and from improving client experience to strengthening institutional capabilities. The industry is shifting away from a one-size-fits-all approach toward a model of human–machine collaboration. As one of the most visible applications of AI in finance, both implementation and ecosystem building have become central concerns.
At the Inclusion·Conference on the Bund wealth forum on September 12, academics, senior fund managers, brokerages and platform executives discussed how AI is evolving from "tool" to "partner," adding value in research, advisory and client services. They also pointed to challenges in user adoption, compliance and risk management, saying openness and collaboration will be critical to progress.
Ant Fortune used the forum to unveil an upgraded version of its wealth management platform, offering three professional AI assistants—for research, operations and content creation—to financial institutions, content creators and other partners. It also expanded full-scenario joint operations to help the industry deliver services more efficiently and improve investor experience.
ZHANG Xiaoyan, vice-dean of Tsinghua University's PBC School of Finance, outlined the evolution of AI in wealth management. Traditional services were mostly person-to-person, with limited reach and uneven quality. In the AI 1.0 era, robo-advisers using machine learning to process structured data cut costs and widened access. Now, in the 2.0 stage powered by generative AI, the focus is on greater professionalism and more human-like interaction.
Generative AI has become adept at handling unstructured data. Zhang cited tools such as iWenCai from Tonghuashun and Wind's Alice, which can screen A-shares with specific valuation metrics in seconds—tasks that once required hours of analyst work.
Human-like qualities are also seen in emotional engagement. She pointed to Ant Fortune's Maxiaocai, which can empathize with anxious investors worried about falling markets before explaining that short-term volatility is normal, shifting AI from a functional tool to a companion.
Zhang also cited academic research. A study led by Peking University's HUANG Yiping found that Chinese users of AI investment tools were mostly young men with higher risk tolerance. These tools not only improved returns but also narrowed capability gaps between investors. In the US, research from California and Georgia showed hedge fund AI usage jumping from 4 per cent to 21 per cent after ChatGPT's release in 2022, lifting excess returns by 4–6 per cent. "Use AI or risk being left behind," she said.
She added that China is ahead globally in applying large AI models in finance, supported by policy and real-world use cases. The State Council's AI+ action plan announced on August 26 provided policy backing, while the wealth management sector offered fertile ground for applications.
LIU Shuoling, CIO of E Fund, said asset management has entered an "AI must-have" stage. Where investment once relied on logic and factor models, it is now a "data war" over time, quality and breadth.
He said the fund hired a geoinformatics PhD to process weather and satellite data, cutting the time needed for AI to predict one-minute stock moves from 13 minutes to just 8 seconds. Active investing is becoming more scientific, he argued, and China's asset managers may even leapfrog into an era of AI-enhanced passive investment strategies. His AI team recently outperformed traditional economists in a GDP forecasting challenge under the "Vision Cup."
WANG Guangxue, executive committee member of CSC Financial, highlighted how AI is reshaping advisory services. With 90,000 licensed advisers serving 200 million stock investors and 700 million fund investors, each adviser handles only 50–100 clients—a bottleneck that compels AI adoption. But issues such as AI "hallucinations" and data privacy still require industry-wide solutions under regulatory oversight.
WANG Jun, president of Ant Group's wealth management division, outlined the firm's open strategy, now in its 3.0 phase. After enabling institutions to manage private domains (1.0) and opening limited public scenarios (2.0), Ant is now offering partners its three AI assistants plus access to 50 billion monthly PVs and trillion-yuan GMV scenarios.
The AI operations assistant can cut content production time by 90 per cent. The research assistant provides real-time monitoring and in-depth report analysis. The content assistant supports the entire cycle from idea generation to diagnostics. Wang stressed that unlike general models, financial AI must prioritise compliance, data accuracy and tool integration to truly reshape the industry.
A roundtable brought together Ant Fortune community content creator JIA Zhi, CSC Financial researcher WU Mahanxu, Maxwealth Fund manager ZHANG Lu and Ant Fortune content operations head YANG Desen.
Jia, who has 660,000 followers, said AI helps manage user interaction by identifying key concerns in hundreds of comments and pre-screening responses for compliance. But he noted AI hallucinations remain a risk and that followers still value "authentic emotional companionship" that machines cannot replicate.
Wu stressed that applying AI in finance requires authority, neutrality and real-time data. AI can track and organize information but must avoid unverifiable or overly absolute conclusions. She added that human intuition—such as linking a robotics conference to investment opportunities—remains indispensable.
Zhang likened AI to a sous chef preparing ingredients: it can summarise lengthy reports, generate stock-picking codes and analyse investor profiles, freeing managers for higher-value tasks. But on-the-ground research, such as factory visits and management talks, cannot be replaced.
Yang said Ant Wealth's AI tools are designed to boost efficiency across roles: enabling retail investors to create content, helping researchers focus on core analysis, and allowing fund managers to provide tailored services.
