Few trust AI for money advice, but many ask it. What US and Canadian advisory firms should do

What the Gallup and Edward Jones survey found

People are asking AI about money, but they do not trust it the way they trust an adviser. Gallup and Edward Jones surveyed 5,075 US adults aged 21 and older (20 March to 6 April 2026) and 2,117 Canadian adults of the same ages (25 March to 3 April 2026), and published the results on 5 August 2026.¹

Among Americans seeking financial guidance, 18% had used AI tools for it.¹ Yet no more than three in ten adults in either country had at least some confidence in AI as a source of financial guidance. Only 3% of Americans and 4% of Canadians had a great deal of confidence in it.¹

Financial advisers sit at the other end. In the same survey, 79% of Americans and 76% of Canadians had at least some confidence in the expertise of financial advisers.¹ One caveat on method: the US sample came from Gallup's probability-based panel, while the Canadian sample was opt-in, so the two countries' figures are not strictly comparable.

Confidence in financial advisers vs AI as a guidance source

US adults: at least some confidence in advisers¹79%

Canadian adults: at least some confidence in advisers¹76%

Either country: at least some confidence in AI (at most)¹≤30%

Low trust, real use: what that means for firms

Low trust does not mean low influence. A prospect who distrusts AI can still use it to shortlist firms, compare fee models or learn what a fee-only planner is. The answer shapes the questions they bring to the first meeting, even if they would never act on it alone.

Our view: for advisory firms, AI is now a research step that sits before the adviser, much like a search engine did fifteen years ago. The prospect asks an assistant, gets a short answer with a few names or a few rules of thumb, and then checks with a person. Firms that are absent from that answer start the relationship a step behind.

The pattern is not unique to consumers. Gartner surveyed 645 B2B buyers in August and September 2025, almost a year before this note, and 69% said they prefer to validate AI-generated insights with sales reps.³ Business buyers use AI, then check it with a person. Retail investors appear to behave the same way, and the adviser is the person they check with.

Why smaller firms are most exposed

Most advisory firms are small, and small firms have the thinnest public footprint for AI engines to draw on. The Investment Adviser Association and Comply count 16,544 SEC-registered investment advisers in their 2026 snapshot of 2025 data (data cut-off not stated), and 92.8% of them employed 100 or fewer people.²

A firm of twelve people rarely has a press office, a Wikipedia entry or a stream of trade coverage. What an engine knows about it comes from its own site, its regulator records and a few directories. If those sources are thin or inconsistent, the answer will be thin or wrong.

The same holds in Canada, where many firms are independent planners and small dealers. We have covered how this works in another regulated market in our piece on financial advice and AI answers in Germany.

What should a firm make citable and verifiable?

Publish the facts a prospect will check, in plain text, on pages an engine can read. Our view: in a low-trust category the winning page is the one that is easiest to verify, not the one with the best slogan.

• Fees and how you are paid. Fee-only, commission or a mix, with the fee schedule or ranges on a page, not in a brochure PDF.
• Services and minimums: who you serve, which account sizes, and what you do not offer.
• Credentials and registration: designations held, and links to the regulator records where a client can confirm them, such as the SEC's adviser search in the US or the provincial and national registries in Canada.
• People: named advisers with short biographies, so an answer about a person resolves to your firm.
• Plain-language explainers on the questions prospects ask first, written by a named adviser and dated.

Then check what the engines say. Ask ChatGPT, Gemini and the others the questions a prospect in your province or state would ask, and ask about your firm by name. Wrong fees, an old address or a retired partner listed as current all happen. Our guide to fixing what ChatGPT gets wrong about your company sets out the order of fixes.

What to watch next

Watch the gap between use and trust. If use keeps rising while confidence stays low, the adviser's role as the person who validates AI answers grows, and firms that publish verifiable facts gain from it. If confidence rises too, the answer itself starts to replace the first meeting for simpler questions, and visibility in that answer matters more.

Either way, the work for a firm is the same this quarter: make the facts citable, check the answers by name in each country you serve, and correct what is wrong at the source.