AI financial advice is surprisingly good — especially if you ask the right questions

Structured ‘academic’ prompts that include detailed financial assumptions significantly improve the quality of AI-generated advice.

Key Points

LLMs generally provide sound financial guidance, steering users toward higher savings, diversified stock funds, and age-appropriate risk-taking, which can lower the barrier to entry for those unable to afford human advisors.
A significant ‘prompt gap’ exists: users with higher financial literacy or prior AI experience generate better advice, leading to wealth disparities (e.g., up to $100,000 difference in retirement wealth at age 60).
AI struggles with dynamic financial planning, such as adjusting to sudden shocks like unemployment or performing active portfolio rebalancing, often relying on overly simplistic rules of thumb.
[AI Synthesis] The tendency of LLMs to recommend specific providers like Vanguard or iShares without user prompting suggests a shift in financial product distribution from search-engine visibility to LLM-driven recommendation moats.