“10주만 더 사고 주식창 닫아라” AI 재무상담, 믿어도 될까

"10주만 더 사고 주식창 닫아라" – AI financial counseling: is it trustworthy?

According to a response provided by ChatPIt, a financial advisor, investors should adjust their portfolio only when their stock allocation goes beyond the target range of 35% to 45%. Users had inputted their investment details into Excel for the AI to analyze. While frequent rebalancing of stock allocation didn't appear to significantly improve risk management, it did increase transaction costs, the AI explained, citing research from U.S. asset manager BlackRock.

During the Korean index fund (ETF) market crash, ChatPIt recommended "mental management" and advised investing 25% in domestic stocks, maintaining a small surplus over the target allocation to support foreign stocks. This illustrates the growing influence of AI in personal financial management. According to a recent OECD policy report on "Artificial Intelligence and Personal Finance," 68% of Korean adults (aged 25 to 59) had used AI for financial-related advice and tasks in 2025.

The report analyzed data from surveys and academic research in major countries, including the U.S., U.K., Canada, and more. Consumers can use AI to input their age, income, investment period, and tolerance for loss to receive asset allocation advice, summarize corporate reports, or compare insurance products. AI can also help consumers understand unfamiliar tax terms and financial products.

The OECD analysis suggests AI can reduce information asymmetry in the finance market and alleviate some of the bias in AI-generated advice. However, AI-generated advice should not be blindly followed. The report warns against the "halo effect," where AI provides seemingly correct but inaccurate or biased responses based on limited or incomplete information.

Additionally, AI may reproduce social biases found in its training data, as demonstrated by a CFI experiment where AI recommended different estate planning steps for male and female respondents. There's also a risk of advertisements or biased recommendations being included in AI responses. The OECD advises users to ask specific questions, verify AI-generated sources, and cross-check critical financial decisions with other data or professional opinions.