AI, Tax Advice, and the Unauthorized Practice of Law
Artificial intelligence is changing how taxpayers obtain answers to legal and tax questions. A taxpayer can now describe a transaction, upload financial information, and ask an AI system what the tax consequences are—all within seconds.
That convenience raises a difficult legal question: When does AI-generated tax information become individualized legal advice, and who is responsible when it does?
The issue is no longer theoretical. Lawyers and tax professionals are already using generative AI for research, drafting, document review, and analysis. Regulators and professional organizations are responding by emphasizing that AI may be a tool, but it does not eliminate the professional obligations that apply to the person using it.
Information Is Not the Same as Advice
The first issue is drawing the line between general tax information and individualized advice.
Explaining what a capital gain is, describing the general requirements for a deduction, or summarizing a provision of the Internal Revenue Code is fundamentally different from telling a particular taxpayer how to structure a transaction to achieve a specific tax result.
The second activity requires applying legal rules to specific facts.That distinction becomes important when AI interacts directly with taxpayers. Consider the difference between asking:
“What is the general rule for deducting business expenses?”
and:
“Here are my business expenses, income, entity structure, and other facts. What should I deduct on my return?”
The second question asks the system to interpret law and apply it to an individual's circumstances. Depending on the jurisdiction and the nature of the service, that can raise questions about whether the activity has moved beyond general information and into regulated professional practice.
Who Is Responsible for the Answer?
AI creates a second problem: accountability.
If a lawyer uses AI to research a tax issue and the AI produces an incorrect citation, the lawyer cannot simply blame the software. The American Bar Association's Formal Opinion 512 emphasizes that lawyers remain subject to their professional obligations when using generative AI, including duties of competence, confidentiality, communication, and candor.
Florida has taken a similar approach. Florida Bar Ethics Opinion 24-1 permits lawyers to use generative AI but makes clear that lawyers remain responsible for their work product and professional judgment. It specifically calls for verification of AI-generated research and emphasizes confidentiality and competence
The principle is straightforward: using AI does not transfer professional responsibility from the lawyer to the machine.
That principle becomes more complicated when there is no lawyer involved at all.
The Unauthorized Practice Question
Unauthorized practice of law rules generally exist to prevent people who are not authorized to practice law from providing legal services that require professional judgment.
AI presents a new version of an old problem.
Suppose a taxpayer enters detailed facts into an AI system and receives a recommendation about whether to make a particular tax election or how to structure a transaction. If the system is effectively analyzing the taxpayer's circumstances and recommending a course of action, the question is no longer simply whether the answer is accurate.
Is the system performing a function that the law reserves for a licensed professional?
There is no single nationwide answer. The regulation of the practice of law is largely state-specific, and the distinction between tax preparation, tax advice, and legal advice can also depend on the circumstances.
That makes sweeping claims about AI and UPL difficult. The relevant question is what the system actually does, who operates it, what professional involvement exists, and what jurisdiction's rules apply.
Accuracy Is Only Part of the Problem
Even if an AI system is not considered to be practicing law, reliability remains a major concern.
Generative AI can produce plausible but incorrect legal analysis, omit important exceptions, or cite authorities that do not support the proposition for which they are offered. Professional guidance therefore emphasizes independent verification rather than treating AI output as authoritative. This is particularly significant in tax law, where a seemingly minor factual distinction can change the result. An answer that sounds confident may still be incomplete.
Where Does This Leave Tax Professionals?
The emerging approach is not to prohibit AI, but to place responsibility on the human professional using it.
For tax lawyers and other practitioners, that means understanding the technology, protecting client information, reviewing AI-generated work, and exercising independent professional judgment. The Florida Bar, for example, specifically requires lawyers using generative AI to consider confidentiality, competence, accuracy, billing, and advertising obligations.
The larger regulatory question is what happens when those safeguards are absent.
AI may eventually perform increasingly sophisticated tax analysis without a human professional involved. If that happens, regulators will have to confront a fundamental question: Should the legal status of an activity depend on whether a human or an algorithm performs it?
For now, the safer distinction is not between “human advice” and “AI advice.” It is between general information and individualized professional judgment. As AI becomes better at crossing that line, tax law and professional regulation will have to determine who is permitted to cross it—and who is accountable when the answer is wrong.
This article is for general informational purposes only and does not constitute legal or tax advice. Unauthorized-practice-of-law rules vary by jurisdiction.
