A new academic study has found that widely available generative AI platforms provide inconsistent, inaccurate, and demographically biased recommendations when answering personal finance questions — raising concerns about the growing number of Americans who rely on the technology for financial guidance.
The research, published last month in the Journal of Financial Planning, examined free-access versions of seven AI platforms: ChatGPT, Claude, Copilot, DeepSeek, Gemini, Meta AI, and Perplexity. The study was authored by finance professors at the University of Georgia and the University of Rome Tor Vergata in Italy — Swarn Chatterjee, Brenda Cude, and Gianni Nicolini.
Researchers queried the platforms in August 2025 using identical prompts across three financial scenarios: emergency savings levels, optimal withdrawal rates from retirement savings, and recommended investment portfolio composition. They then reissued the same prompts while altering the race and gender of the hypothetical individual, testing whether recommendations would shift based on demographic variables.
The results showed "substantial variation in guidance" across platforms on emergency savings and asset allocation. "Although the tools often produced recommendations that broadly aligned with generic financial planning principles, such as the 4 percent retirement withdrawal rule, there were significant differences across platforms in suggested emergency savings and portfolio allocations," the researchers wrote.
The study's authors concluded that AI's "suboptimal" or biased outputs raise questions "about the consistency and fairness of GenAI-driven recommendations." They added that "GenAI-driven responses may sound confident but can still be incomplete, misleading, or incorrect."
The findings arrive as AI use for financial guidance has become widespread. Two out of three Americans — 66% — who have used generative AI said they have leveraged it for financial advice, according to an Intuit Credit Karma survey published in September. Among Gen Z and millennial users, that share rises to 82% for each cohort.
Experts say AI performs adequately when providing high-level overviews — explaining the general case for investment diversification, for instance, or comparing exchange-traded funds to mutual funds in broad terms. The limitations become more acute when the questions turn specific.
"One of the things about LLMs that I find particularly concerning is that no matter what you ask it, it'll always come back with an answer that sounds authoritative, even if it's not," said Andrew Lo, director of MIT's Laboratory for Financial Engineering and principal investigator at its Computer Science and Artificial Intelligence Lab, in an interview in March. "When it comes to very, very specific calculations of your own personal situation, that's where you have to be very, very careful," Lo added.
Analysts and researchers point to several structural reasons for AI's limitations in this domain. The platforms are sensitive to prompt phrasing, meaning small differences in how a question is worded can produce meaningfully different recommendations. AI tools also do not carry a fiduciary duty to users, meaning they have no legal obligation to provide advice in users' best interests — a standard that licensed financial advisors must meet.
The phenomenon of AI "hallucination" — in which large language models generate plausible-sounding but factually incorrect information — compounds the risk for users who accept AI outputs without scrutiny.
The latest study's conclusions echo those of earlier research. A 2024 study examining ChatGPT's financial advice capabilities, published in the Journal of Risk and Financial Management, found the tool's recommendations to be "generic" and prone to overlooking pertinent details. "We believe that ChatGPT can serve as a starting point in giving and finding financial advice, but its recommendations should be carefully scrutinized and assessed," those researchers wrote.
The Journal of Financial Planning study's authors acknowledged that generative AI tools "are still evolving," and noted that outputs from paid versions of the platforms may differ from the free-access versions they assessed. Their overall recommendation: treat AI as a complement to professional financial advice, not a replacement for it.
Disclaimer