The expansion of artificial intelligence (AI) wealth management services has been driven by the rapid migration of financial customers to non-face-to-face channels. Growing trust stemming from widespread use of generative AI tools such as ChatGPT and Gemini has also played a role. Commercial banks, viewing AI services as a way to absorb retail investment demand, are preparing strategies accordingly.
KB Kookmin Bank's "K-Bot Ssaem" has demonstrated the potential of AI wealth management. When this reporter requested investment advice for 1 million won through the "AI Portfolio" feature on the KB Star Banking app on the 20th, the system proposed a portfolio composed of 91% stocks, 4% bonds, and 5% alternative assets. Linking MyData from other financial institutions, completing an investment propensity analysis, and actually subscribing to the product took less than 10 minutes.
The completed portfolio had a heavy weighting of U.S. information technology (IT) stocks, including Nvidia, Alphabet, and Tesla. The AI explained that, based on a comprehensive analysis of historical data by country and asset class, fund performance, and fund flow trends, U.S. equities were the asset with the highest investment preference.
By contrast, the "customized portfolio" recommended by a human expert consisted of 25% bonds and 75% stocks. On the equity side, domestic and overseas aerospace and defense stocks accounted for a large share.
While the sign-up time was similar for both approaches, the AI portfolio's competitiveness stood out more clearly in the non-face-to-face channel. The AI portfolio enables personalized recommendations that incorporate assets held at other financial institutions. It also breaks asset allocation into 11 categories — including U.S., European, Japanese, and Chinese equities — compared with the five categories used by the expert portfolio (domestic stocks, emerging stocks, developed stocks, domestic bonds, and overseas bonds), offering a more granular asset allocation. The ability to receive a more detailed recommendation in a short time is seen as drawing customers accustomed to non-face-to-face investing toward AI.
This trend is also reflected in the growth of the banking sector's AI-based robo-advisor market. As of the end of March, based on companies that had passed the Koscom test bed, the number of bank robo-advisor clients stood at 177,659, up 5.5% from 168,335 a year earlier. Assets under management reached 642.5 billion won ($470 million), up 5.6% from 608.6 billion won a year earlier.
A simple calculation puts the average amount managed per person at about 3.61 million won. While differences in individual investment size must be taken into account, the figures suggest considerable demand for entrusting small amounts of money through non-face-to-face channels rather than managing large asset pools. "Robo-advisor products are mainly targeted at office workers who do not have time to check the market every day to pick funds or stocks themselves," an official at a commercial bank said.
However, commercial banks say that high-net-worth investors still prefer face-to-face service. Although the number of bank branches has been declining amid accelerating AI adoption, the number of private banking (PB) centers at the four major banks — KB, Shinhan, Hana, and Woori — rose from 74 at the end of 2021 to 81 at the end of March this year. KB Kookmin Bank increased its PB workforce by 34% over five years, from 77 in 2021 to 103 as of February this year.
This is because wealthier clients still prefer face-to-face consultations with PBs. In a survey of 713 wealthy individuals conducted last year by the Hana Institute of Finance, only 18% said they could trust financial transactions conducted through AI recommendations or entrusted to AI. In a survey on preferred methods of wealth management consultation, the most common response, at 38%, was "being managed primarily by staff while using AI when necessary." Only 6% said they wanted management handled solely by AI. For high-value investments, demand remains strong for referring to AI while relying on human experts for final decisions.
"In areas that carry high risk and are highly personal, such as finance, customers tend to trust and rely on people more than AI," the KB Management Research Institute said. "The presence of human experts serves as a safety zone that eases customer anxiety and enhances trust in the overall system." A PB at a commercial bank added, "From the staff's perspective, we can predict which products a client will prefer or avoid based on the client's tendencies that we have identified through a long consultation process."
