Asking an AI before an accountant or lawyer: 13% would, 1% would go straight to the professional, and most would search or go it alone
13.0% of US adults would ask a free AI assistant first with a tax question an accountant would charge $200 to answer, against 1.0% who would go straight to the accountant (n=200, September 2026); 86.0% would search online, ask family or work it out themselves.
- Who was asked
- US adults, 200 clones, US
- When
- September 2026 (fielded 25 September 2026)
- How
- Predicted responses from Brox clones, each built from a long-form interview with a real person.
- What changed from the obvious question
- The obvious question is 'would you use AI instead of an accountant or lawyer?', which invites a two-way answer. This study priced each professional, wrote a real situation for each, and offered search, family and doing it yourself alongside AI, because those are what people actually do first.
Section 01: What the panel said
- #
13%
13.0% would ask a free AI assistant first with a $200 tax question, and 1.0% would go straight to an accountant; 86.0% would search online, ask family or work it out themselves.
base 200
- #
2
2 of 200 people would go straight to a paid professional in any of three situations (tax, a deposit dispute and a side business); a free AI assistant was the first move for 13.0%, 5.5% and 7.0% respectively.
base 200 · contrast
- #
55.5%
55.5% say business planning is where they would be most comfortable relying on AI alone, yet only 14 of those 111 would start with an AI when a $500 side-business decision is put to them.
base 200 · contrast
- #
61%
61.0% say high stakes (a lot of money or a legal deadline) would most make them still pay a professional after an AI had answered; 3.0% say nothing would.
base 200 · caveat
13.0% would ask a free AI assistant first with a $200 tax question, and 1.0% would go straight to an accountant; 86.0% would search online, ask family or work it out themselves.
02Who is the paid advisor really competing with for the first move?
Not mainly AI. Across three situations, a $200 tax question, a $1,500 deposit dispute that a lawyer would charge $300 to discuss, and a $500 planning session for a side business, 2 of 600 first moves went straight to a paid professional. The rest went to a search engine, to family or friends, or to working it out alone: 86.0% for the tax question, 94.5% for the deposit and 93.0% for the side business (n=200 each).
A free AI assistant took a small slice: 13.0% of first moves on the tax question, 5.5% on the deposit and 7.0% on the side business. Across all three, 26 of 200 people would start with an AI at least once.
03Why do people name business planning as safe for AI and then not use it?
Asked where they would be most comfortable relying on a free AI assistant alone, 55.5% chose a business or career planning question, 4.0% a tax question and nobody a legal one; 40.5% chose none of these (n=200). Yet of the 111 who picked business planning, only 14 would start with an AI when the $500 side-business decision was put to them. The other 97 would search, ask family or do it themselves.
The age split is wide. 74.7% of 18 to 44 year olds named business planning (n=91), against 39.4% of those aged 45 and over (n=109), and 57.8% of the older group would rely on AI alone for none of the three.
04What still sends people to a paid professional?
Stakes, above all. If an AI had already answered, 61.0% said a lot of money or a legal deadline being at stake would most make them still pay a professional, and 28.5% an AI answer that was unclear or seemed wrong (n=200). Only 6.5% named accountability and 1.0% a document that has to be prepared, signed or filed. Just 3.0% said nothing would.
05What would an accounting, legal or advisory firm do with this?
- A reading, not the panel's answer: the free first step people take is research, so a firm that publishes the plain answer to the common question is present at the moment most people start, whether they search or ask an AI.
- A second reading: the fee is the obstacle people describe again and again in their reasons, so an entry point priced for a single question may matter more than any argument against AI.
- A third: people come back when the stakes rise or the free answer wobbles, so the message that lands is about the cost of getting a high-stakes answer wrong, not about AI being unreliable in general.
Section 06: What they said, and the interview line it rests on
“Since we're retired and living on a limited income, we have to make our money stretch, so I certainly wouldn't want to spend $200 right off the bat if I don't have to. Whenever something like that comes up, my normal reaction is to do a little research first and see what my options are. I would search online or ask friends or family who might have dealt with a similar tax situation. If I still couldn't figure it out or if it was truly complicated and important, then I might consider consulting a professional, but searching online and checking with people I know would be my first step.”
Predicted from this line in the interview
“Now that we're retired on limited income, we have to make it stretch a lot further.”
“I would definitely ask a free AI assistant and act on its answer. Chatbots and AI tools are advancing rapidly and becoming more normalized and capable every single day, so there's no reason not to take advantage of them—especially when they're free. I'm working from the internet and striving to build up capital for my business, so dropping $200 on an accountant for a question an AI can solve makes zero sense to me.”
Predicted from this line in the interview
“However, I think the humanoid robots and chatbots and all the other AI factors will begin to be more normalized and better interactions to the human.”
“Honestly, I'd definitely search online or talk to my family first. A $500 fee for a business consultant is way too expensive for me, especially right now when I'm trying to be careful and save money. Whenever I need advice or have to make a big decision about money or anything important, I always go straight to my family—my mom, my dad, and my sisters. They're the people I trust most, so I would definitely ask them what they think before doing anything else or spending that kind of money.”
Predicted from this line in the interview
“and I would look for advice and my mom were really just my family my mom my dad and my two sisters they definitely would help me with advice and influence my choice for sure but yeah I would save give it to my mom and that's who I look for advice to”
Section 07: How much should you trust this?
What we would defend
- Almost nobody makes a paid professional their first move at these fee levels: 2 of 600 scenario answers, stable across two pulls of the data.
- Search, family and do-it-yourself dominate the first move in all three situations (86.0% to 94.5%, n=200 each), and a read of the 287 search-or-family answers found that the few that mention AI mostly reject it, so that option is not hiding AI use.
- High stakes are the main reason to still pay a professional (61.0%, n=200).
- The age gap on business planning (74.7% of 18-44, n=91, against 39.4% of 45 and over, n=109).
What we would not
- The split between acting on the AI answer and having it checked (19 against 7 on the tax question): counts too small to read as a pattern.
- That the profession alone drives the differences between scenarios: fee, stakes and profession all change together across the three situations.
- The 'check with a professional' option in the deposit and side-business situations: Ask Brox fielded the three as one question with one answer list, so it read 'then check with an accountant' there (0 and 6 answers).
- That people never pay a professional in the end: the question asked only about the first step.
- That 55.5% trust AI with business planning in an absolute sense: the question asked which was most comfortable, with 'none of these' available.
The main limitation
The study asks what people would do first in three written situations, not what they did, and the three situations change the fee, the stakes and the profession at the same time.
Section 08: How was this run?
| Segment | Clones |
|---|---|
| Aged 18-29 | 43 |
| Aged 30-44 | 48 |
| Aged 45-64 | 66 |
| Aged 65 and over | 43 |
Question wording as fielded
- What would you do first? (three scenarios fielded as one question: a tax question an accountant would charge $200 to answer; a landlord keeping a $1,500 security deposit, with a lawyer charging $300 for a one-hour consultation; deciding whether to start a small side business, with a business consultant charging $500 for a planning session)
- If a free AI assistant had already answered your tax, legal or business question, which ONE of these would most make you still pay a professional?
- For which ONE of these would you be most comfortable relying on a free AI assistant alone, without paying a professional?
What was not asked, and why
Past use of AI for tax, legal or business questions was not asked, because the panel does not hold that behavior as data. Whether people eventually paid a professional was not asked; only the first step. No firm or AI brand was named.
What changed from the obvious question
The obvious question is 'would you use AI instead of an accountant or lawyer?', which invites a two-way answer. This study priced each professional, wrote a real situation for each, and offered search, family and doing it yourself alongside AI, because those are what people actually do first.
Fielded through Ask Brox on 25 September 2026.
More on AI or an adviser? · Consulting & professional services · U.S. nat. rep. (US) panel
Ask this of your own customers: humans@brox.ai