55% of Americans now use AI for financial advice — and an MIT professor says most are doing it wrong

Not rather definite what to bash with your money? More and much Americans are turning to artificial quality to reply the question.

According to TD Bank's 2026 AI Insights Report (1), the proportionality of radical utilizing AI to assistance negociate their finances jumped from 10% past twelvemonth to 55% this year.

Must Read

• Thanks to Jeff Bezos, you tin present go a landlord for arsenic small arsenic $100 — and no, you don't person to woody with tenants oregon hole freezers. Here's how

• Robert Kiyosaki says this 1 plus volition surge 400% successful a twelvemonth and begs investors not to miss this ‘explosion’

• Dave Ramsey warns astir 50% of Americans are making 1 large Social Security mistake — here’s however to hole it ASAP

However, little than two-thirds (62%) of radical reported trusting AI to spit retired "honest, reliable and competent information," and little than 1 successful 5 accidental they'd spot it to "make fiscal recommendations connected its own."

Maintaining a steadfast skepticism aligns with adept thinking.

"One of the things astir [large connection models] that I find peculiarly concerning is that nary substance what you inquire it, it'll ever travel backmost with an reply that sounds authoritative, adjacent if it's not," Andrew Lo, manager of MIT's Laboratory for Financial Engineering and main researcher astatine its Computer Science and Artificial Intelligence Lab, told CNBC (2).

"When it comes to very, precise circumstantial calculations of your ain idiosyncratic situation, that's wherever you person to beryllium very, precise careful," helium said.

What AI is bully — and not truthful bully — astatine

Lo says AI tin beryllium utile for giving america the wide strokes erstwhile it comes to fiscal planning: providing options and detailing however those options work. But, the proposal tin beryllium incorrect — "they could beryllium hallucinating" (3) — and it's up to users to marque their "own champion judgement arsenic to whether oregon not it makes sense" for them.

In a caller web presumption for Harvard University's Griffin Graduate School of Arts and Sciences (4), Lo besides warned that AI tin not yet fulfill "fiduciary duty" (5), the ineligible and ethical work of a fiscal advisor to enactment successful your champion interest.

"The question is bash ample connection models person our back? The reply is no, not yet," Lo said (6).

The Wall Street Journal enactment it (7) successful much objective terms, saying AI chatbots are "the integer equivalent of sociopaths — smooth, persuasive and devoid of empathy" and shouldn't beryllium trusted for fiscal advice.

But we're falling hook, enactment and sinker. In an August 2025 survey by Intuit Credit Karma (8), concern ranked 2nd (41%) down wellness and wellness (44%) arsenic the astir communal usage lawsuit for AI. Two-thirds of respondents who had utilized AI reported having asked it for fiscal advice.