How Can Service Firms Deal With Clients Who Consult ChatGPT First?

The old model of business, where a client comes to you because you hold relevant information and can take action to resolve their issue, is being disrupted by clients who come in with half-baked information from AI and then expect businesses to implement it. Not surprisingly, the answers generated by AI can be wrong, as they don’t have access to real-life context or sufficient up-to-date information on the issue itself. Tell the client that they’re wrong, and you risk offending them. Proceed with the AI’s conclusion, and you risk someone’s savings.

This is happening across all industries, from patients arriving at hospitals with AI-generated diagnoses to travelers challenging travel agents with their AI-made custom itinerary to destinations that don’t even exist. The era of AI-briefed clients is presenting a challenge to many industries. How does accounting deal with this issue?

The AI-Briefed Client

Imagine Kayla. She runs a small business in Queens. One day, she forwards her accountant a screenshot from ChatGPT where the chatbot is advising her to change how her business is registered in order to save thousands in taxes every year. Theoretically, it works, but not for her business. Local NYC rules and the salary she’d be required to pay herself would cancel out most of the savings. Her accountant has to explain all this to her without sounding defensive so that Kayla doesn’t make an expensive mistake. However, some people in this case might start to distrust their accountant because they are being presented with an answer by a chatbot that seems superficially accurate.

Such cases are becoming common now. OpenAI reported that tax-related ChatGPT queries in the U.S. quadrupled from Q1 2025 to Q1 2026. Its own chart carried a disclaimer that ChatGPT isn’t meant to replace professional advice. Similarly, Adobe found that 26% of U.S. workers planned to use AI for tax filing in 2026, up from 11% in 2024. This means clients seem to be superficially informed and the onus of checking their answers falls on the professional.

Why Accounting Is Particularly Exposed to This

While taxation rules can be accessed from anywhere, their application varies according to the client’s location, how the business is registered, the tax year, the client’s history, and other things outside the model’s training data. Moreover, things like thresholds, deadlines, and credits shift annually and a model might apply an outdated rule. TaxBench’s 2026 evaluation of 16 frontier models showed that the best model answered tax-knowledge questions correctly 84% of the time on a first attempt, but only 42% of the time across five attempts on the same question.

The hidden cost of these mistakes falls on the professional who has to find the fallacy in the AI’s reasoning, explain it to the client, and argue for why the AI is wrong. Correcting a client can make the service provider look defensive or out of date, creating a trust paradox; the client sees a perfectly mapped-out plan given to them by a chatbot and wonders if they’re being lied to. Most firms don’t even charge for this sort of work. But the liability stays with the firm nevertheless. Coupled with the shortage of experienced accountants, this additional workload might be the proverbial straw that breaks the camel’s back.

Practical Guidance for Firms

With firms already struggling with enormous workloads, a strategy must be put in place to deal with this new challenge.

1. Clarify expectations when onboarding the client. Explain how the firm treats AI-generated advice and who is accountable for what gets filed. This clears the gray area where work gets put in but isn’t billed or a practitioner takes the fall for a decision that wasn’t theirs to begin with.
2. Use clients’ AI answers as input on what their real concerns are. Ask the client to share what they asked and what they got back to trace the questions that might be bothering them but they’re afraid to ask.
3. Share how you verified your conclusion. Point to the publication or rule that guided your decision-making instead of being dismissive of their AI-generated strategy. This helps them to understand the shortcomings of the chatbot as well.
4. Build a library of recurring misconceptions. An FAQ, if you will. The same wrong answers keep coming back, so standardize the responses and share them in advance.
5. Lastly, deploy AI in your own operations to free up your accountants’ time to have these conversations. Fully manual firms will struggle to implement any of the above solutions because they will simply be too overwhelmed with work to also try to answer extra questions.

The Impact on Service Firms

Information becomes more abundant each day as technology progresses. Service firms are no longer gatekeepers of knowledge as theoretically, everyone can find out how to acquire the service on their own. The selling point then is time and peace of mind. Accounting acts as a trust layer in financial infrastructure. As clients increasingly arrive with a pre-built plan from a chatbot, the professional’s value moves from explaining to verifying and deciding. The same holds for law firms, consultancies, and advisory practices. That trust can only be built by people.

The test is whether firms are freeing up their accountants’ time to deal with this sort of relationship work. While the industry struggles with an overall dearth of accountants, AI actually provides a solution by automating routine work. So AI-briefed clients are in fact heralding a change in the industry to adopt AI to ease the burden of their workloads, while leaving the actual decision-making and problem-solving to the professionals. The client will always have a second opinion now. The firm’s job is to be the one they trust.