Growth Minded Accountant Podcast

AI Knows Tax Law. But Does It Know Your Client?

AI is getting remarkably good at tax research, document analysis, summarization, and identifying patterns.

But knowing the tax law is only part of the equation.

The harder question is whether AI understands enough about the individual client to recognize what actually matters.

In this episode of The Growth Minded Accountant, Lee Reams is joined by Mike Gleeson, Director of Tax Intelligence at CountingWorks PRO, to explore the next evolution of AI for tax and accounting firms: client intelligence.

A tax answer can change based on seemingly small differences in a client’s facts. The same is true for advisory opportunities. A business sale, retirement decision, real estate transaction, income change, financing need, or unresolved conversation may completely change what the professional should be thinking about.

That is where context becomes critical.

Lee and Mike discuss how modern AI systems can combine tax intelligence with client history, documents, prior conversations, financial information, and other signals to help tax professionals recognize opportunities that might otherwise remain buried inside the firm.

They also explore why AI-generated information should not all be treated the same. Firms need to distinguish between verified facts, estimates, and AI-inferred information, while maintaining appropriate consent, professional review, and human judgment.

The goal is not autonomous tax advice.

It is better preparation.

Better context.

Better questions.

And ultimately, better decisions.

Because AI may know the law.

Client intelligence helps it understand the situation.

And the professional still decides what to do.

In This Episode, You'll Learn:

Key Takeaways

1. Tax intelligence without client intelligence is incomplete

AI may be able to research a tax rule accurately, but the correct application depends on the client’s actual facts.

The more context the system understands about the client, the more useful it can become to the professional.

2. Nuance matters enormously in tax

A seemingly small factual difference can change the answer.

Lee and Mike discuss why firms should not assume that a generic AI response is automatically correct simply because it sounds confident. Professional review, reliable sources, and specialized tax intelligence remain essential.

3. Client context can turn AI from reactive to proactive

Traditional systems often wait for the accountant or client to ask the question.

Client intelligence can help recognize signals earlier: a retirement approaching, a business transition, an income change, a real estate event, a financing need, or a previous issue that was never resolved.

Instead of only answering questions, AI can help professionals identify which questions deserve to be asked.

4. Existing client data contains hidden advisory opportunities

Tax returns, client conversations, intake information, documents, prior recommendations, and financial activity can collectively reveal opportunities that may be difficult for a human to monitor across hundreds of relationships.

AI can help surface those signals so the professional can decide whether a conversation is warranted.

5. AI can help smaller firms operate with deeper institutional knowledge

Historically, much of a firm’s client knowledge lived in the memory of senior partners.

Client intelligence can help capture that institutional knowledge and make relevant context available to the broader team, reducing dependence on one person remembering every detail about every relationship.

6. Personalized intake can become far more useful than a generic organizer

Once AI understands a client’s history, firms can move toward individualized intake questions based on prior returns, conversations, unresolved issues, and known changes.

That creates the potential for better information before the professional even begins the engagement.

7. Not all AI-generated information should be treated as fact

One of the most important distinctions in the episode is separating:

  • Verified information
  • Estimated information
  • AI-inferred information

Professionals should know which type of information they are reviewing and why the system reached a particular conclusion.

8. Consent and data controls matter

Client intelligence becomes more powerful as more information is connected, but firms also need to understand when consent is required, particularly when tax return information may be used or disclosed beyond the original engagement.

Technology should make these controls easier, not obscure them.

9. Human judgment remains the final layer

The model discussed throughout the episode is not “AI decides.”

It is:

AI understands.
AI prepares.
AI surfaces the signal.
The professional decides.

That distinction becomes even more important as AI systems become more capable.

10. The real AI advantage may be context, not simply model intelligence

Firms should not only ask, “How smart is this AI?”

They should also ask:

Does it understand what changed?

Does it remember what we discussed?

Does it know what was left unresolved?

Can it distinguish fact from inference?

Can it recognize when something deserves professional attention?

That is where client intelligence begins to create a much more valuable system.

Transcript

We know that everyone digests information differently. That’s why we’re now sharing the full transcript of each episode of The Growth Minded Accountant right here on the CountingWorks PRO blog. Whether you’re short on time, like to scan and highlight, or simply prefer reading over listening, you can catch up on every conversation at your own pace.

Each week, we cover topics that matter most to tax and accounting professionals—from AI and automation to marketing strategies, firm growth, and client relationships. Scroll down to read the full episode, or subscribe to the podcast to listen on the go.

Frequently Asked Questions

Q. What is client intelligence for a tax and accounting firm?

Client intelligence is the ability to bring together information about a client’s tax situation, financial history, documents, conversations, business activity, previous recommendations, and other relevant context so the professional can better understand what the client may need next.

The goal is not simply to store more data. It is to make that information useful.

Q. Why isn’t tax knowledge alone enough for AI?

Tax rules are highly dependent on facts.

Two clients asking the same question may receive different answers because of income, filing status, ownership structure, timing, dependents, business activity, prior elections, or other circumstances.

AI needs both reliable tax intelligence and accurate client context to provide useful professional support.

Q. Can I trust AI to answer tax questions?

AI can be extremely useful for tax research and analysis, but professionals should not assume every answer from a generic chatbot is correct.

Tax law changes, nuance matters, and AI systems can misunderstand or omit important facts.

Professional review, references, current tax sources, and specialized tax systems are still important.

Q. How can client intelligence help identify advisory opportunities?

Client intelligence can help recognize changes or signals that may deserve a professional conversation.

Examples could include retirement, a business sale, real estate activity, financing needs, income changes, estate planning issues, business growth, or a previously discussed strategy that was never completed.

AI can surface the signal. The professional determines whether it matters.

Q. Does client intelligence replace the accountant’s judgment?

No.

The model discussed in this episode is specifically designed to keep the professional in control.

AI can gather information, organize facts, identify patterns, prepare research, and surface opportunities.

The tax or accounting professional still determines what advice should be given and what action should be taken.

Q. What is the difference between verified, estimated, and AI-inferred information?

Verified information is supported directly by a reliable source, such as a tax return, client document, or confirmed client response.

Estimated information is calculated or approximated based on available data.

AI-inferred information is a conclusion or possibility the system has identified based on patterns and context.

Those categories should be clearly identified so the professional understands what they are reviewing.

Q. Can AI remember previous client conversations?

When client information and communication history are connected to the intelligence system, AI can potentially use prior conversations as context.

That can help the professional remember previous recommendations, unresolved questions, client concerns, or planned follow-up rather than starting from zero each time.

Q. How could AI improve the client intake process?

Instead of sending every client the same generic organizer, AI can help generate questions based on the client’s actual history.

For example, it may ask about something that appeared on last year’s return, a business change discussed during a prior meeting, or an issue that remained unresolved.

The result can be a more relevant and efficient intake process.

Q. What role does client consent play in client intelligence?

Consent can be important when client or tax return information is used or disclosed outside the purpose for which it was originally collected.

Tax professionals need appropriate processes for handling client information and, when necessary, obtaining the required authorization or consent before information is shared or used in other ways.

Q. Can client intelligence help smaller accounting firms compete with larger firms?

Potentially, yes.

A smaller firm may not have multiple senior partners who collectively remember every detail about every client.

AI-assisted client intelligence can help organize that knowledge and make relevant context easier for the professional and staff to access.

That can allow boutique firms to deliver a deeper, more consistent client experience without trying to replicate a large-firm staffing model.

Q. What should firms look for when evaluating AI for tax work?

Do not evaluate the system only on how impressive the chatbot appears.

Ask whether it provides reliable sources, whether its tax knowledge is current, how it handles client context, whether it distinguishes verified facts from inferences, how consent is managed, and whether the professional remains in control of the final decision.

Related Episodes & Resources

Listen to other podcast episodes or read other related blog articles with relevant information and insights.