Do Different Generative Artificial Intelligence (GenAI) Tools Provide Different Financial Recommendations? | Financial Planning Association
Skip to main content
Accessibility
FPA-RGB
Close search overlay
Trending<br>Membership Benefits<br>Pro Bono<br>Continuing Education<br>Annual Conference
Looking for more FPA resources?
FPA Learning Center<br>Live Webinars and On-Demand Programs
FPA Connect<br>Compliance-Friendly Online Community
Do Different Generative Artificial Intelligence (GenAI) Tools Provide Different Financial Recommendations?
By
Gianni Nicolini, Ph.D.; Brenda J. Cude, Ph.D.; and Swarn Chatterjee, Ph.D.
June 2026
Journal of Financial Planning: June 2026
Audio file
NOTE: Please be aware that the audio version, created with Amazon Polly, may contain mispronunciations.
Executive Summary<br>This study evaluated the consistency and fairness of financial recommendations generated by seven widely used GenAI tools across three core areas of household financial planning: emergency savings, retirement withdrawal rates, and investment portfolio allocations.<br>Prompts describing standardized case scenarios were presented to each tool, with the race and gender of the household lead varied in subsequent entries to assess potential demographic bias in recommendations. Recommendations were recorded and analyzed using ANOVA and Tukey’s HSD tests.<br>Results showed substantial variation in guidance across platforms. Emergency savings recommendations ranged widely, and these differences were statistically significant.<br>Retirement withdrawal recommendations were more uniform, with most tools’ recommendations aligning with the traditional 4 percent rule.<br>Portfolio guidance varied significantly, particularly in equity, cash, and alternative asset allocations.<br>Some tools provided consistent recommendations across demographic groups, whereas others produced differing outputs, indicating potential bias.<br>Overall, GenAI tool recommendations generally aligned with conventional financial principles but varied meaningfully in specificity and consistency. The findings suggest that GenAI may serve as a helpful starting point for consumers but should complement, not replace, professional financial advice. Continued attention from practitioners and researchers is essential to ensure transparency, fairness, and accuracy in GenAI-driven financial guidance.
Dr. Gianni Nicolini is a professor in the area of banking and finance in the department of management and law at the University of Rome Tor Vergata, Italy (https://web.uniroma2.it/en). Dr. Nicolini’s main research interests are financial literacy and financial education and their relevance in financial consumers’ decision-making process.<br>Dr. Brenda J. Cude is a professor emeritus in the department of financial planning, housing and consumer economics at the University of Georgia (www.uga.edu). Dr. Cude is an internationally renowned scholar in financial literacy.<br>Dr. Swarn Chatterjee, CFP®, is a professor in the department of financial planning, housing and consumer economics at the University of Georgia. Dr. Chatterjee conducts research that examines financial planning performance, the link between financial well-being and health across populations, and factors that enhance individual financial decision-making.<br>NOTE: Click on the images below for PDF versions.
Because generative artificial intelligence1 (GenAI) is relatively new and currently evolving, there is little research about its use to provide financial recommendations to consumers. In a 2025 survey of 1,326 U.S. adults with household incomes between $50,000 and $200,000, more than 50 percent indicated that the internet was their top source for financial guidance (Fischer 2025). If these searches do not yet involve using GenAI tools, they will soon (Capgemini 2025).<br>Financial planners are increasingly encountering clients who arrive after consulting GenAI tools, often unaware of their limitations or potential biases. For example, GenAI tools may rely on incomplete or outdated information. They may lag in updating important tax-related data, such as 401(k) contribution limits, standard deductions, and catch-up contribution limits. This study provides the first systematic comparison of multiple GenAI platforms across core planning domains while also evaluating whether demographic characteristics influence the recommendations provided.<br>Variability and potential inequities in GenAI-based financial planning recommendations may arise from several structural and technical factors. GenAI systems are trained on large datasets that may not equally represent different demographic groups, income levels, or financial behaviors, which can lead to recommendations that are more suitable for some users than others. Historical financial data may also reflect longstanding inequalities in access to credit, investment opportunities, and wealth accumulation, patterns that may be unintentionally reproduced in GenAI-generated outputs (Shabsigh and...