Original title: AI Financial Advice: Supply, Demand, and Life Cycle Implications
Taha Choukhmane, Tim de Silva, Weidong Lin, Matthew Akuzawa · August 2026
People often struggle with how to spend and save their money over a lifetime. To see if artificial intelligence can help, researchers asked a representative group of people to write questions asking an AI for spending and investing advice. They tested this using GPT-5.2 and simulated what would happen over the people's lives under realistic job and market conditions. They found that following the AI's guidance moves people toward established economic theory. It suggests owning more diversified stocks, holding fewer stocks as you age, and building bigger savings cushions. However, the advice changes based on who asks. Recommendations shift by gender, financial knowledge, and AI experience. Men and women get different stock advice partly because they ask differently and partly because of gender labels.
Imagine asking two friends for directions to the same restaurant. One asks 'Where is it?' and the other asks 'Which route is safest?' — the two answers differ because the questions differ. Then the friend hears the asker's gender and changes the directions even when the wording is identical. In this paper, the 'friend' is an AI adviser; the wording is the written request, and a gender label attached to that request shifts the recommended mix of stocks and savings.
AI financial advice can affect how much people save and how they invest over a lifetime. The paper shows the advice is not neutral: it changes based on who asks, how they ask, and even a gender label on an identical request. Similar people can therefore be steered toward different portfolios, which matters for anyone relying on an AI for financial decisions.
The abstract doesn't say whether these results generalize to other LLMs or to real people actually following the advice, nor does it describe the direction of variation by financial literacy or AI experience. It does cover the method and the simulated GPT-5.2 results, but not the size of the portfolio changes in dollars.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
AI Financial Advice: Supply, Demand, and Life Cycle Implications — Taha Choukhmane, Tim de Silva, Weidong Lin et al.
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.