Blogs

Better late than never – Guidance on responsible AI use for tax professionals

7/2/26

By: Will Covino and Nancy Reimer

The IRS Office of Professional Responsibility recently issued its Introductory Guidelines for Responsible AI Use in Federal Tax Practice—a long overdue acknowledgment that artificial intelligence is now embedded in the day-to-day work of tax professionals. At one level, the guidance feels new. At another, it is not new at all. That is the point.

The IRS appropriately recognizes what practitioners already know: generative AI is both transformative and risky. On one hand, it offers cost savings, rapid data analysis, and improved workflow efficiency. On the other hand, it introduces very real problems: fabricated outputs (often referred to as “hallucinations”), bias, lack of transparency, and confidentiality risks. Those risks are not theoretical. Courts have already sanctioned attorneys for relying on unsupported AI-generated content. The IRS is signaling that AI is a tool to be used with an understanding of its limitations and with appropriate safeguards in place. It is not a shield or an invitation to disregard the fundamental duties and obligations imposed by IRS Circular No. 230.

The IRS emphasized several key principles.

• Due diligence. Practitioners using AI are expected to “thoroughly review all AI-created documents and language incorporated into writings before delivery to a client or submission to the IRS.” They should independently verify facts, confirm citations, and validate conclusions. Put simply, AI is only a starting point. It is intended to augment a professional’s judgment, not replace it. That principle is not new, but it takes on greater importance where outputs can appear polished and authoritative while still being incorrect.
• Competence. It is not sufficient for practitioners to simply understand the tax code—practitioners must also understand the tools they are using. This includes recognizing how AI generates outputs, identifying where it may be unreliable, and exercising judgment in deciding when not to rely on it. Blind reliance is not just risky; it is increasingly difficult to reconcile with the duty of competence itself.
• Billing. Circular 230’s prohibition on unconscionable fees still applies. The fact that AI can significantly reduce the time required to complete work does not permit practitioners to bill as though that efficiency does not exist or to double-bill for AI-assisted tasks
• Confidentiality. AI systems, particularly those provided by third parties, raise legitimate concerns about how client information is used, stored, and potentially repurposed. The guidance highlights the risk of sensitive taxpayer information being entered into unsecured or public systems. Practitioners must therefore be deliberate in how client information is handled, ensuring that appropriate safeguards are in place and that any tools used maintain data within secure, controlled environments to prevent unauthorized disclosure.
• Responsibility. The IRS further emphasizes that AI is not merely an individual responsibility—it is a firm responsibility. Firms must deploy internal policies and procedures for compliance with IRS Circular No. 230. This should include training and procedures addressing how AI is used, how outputs are reviewed, and how data is protected. These expectations tie directly back to existing supervisory obligations, and in practice, firms using AI without structure or oversight are exposing themselves to unnecessary risk.

The overarching message is simple: the practice of public accounting remains guided by the same core principles, irrespective of the convenience new technology may bring. AI may change how work is produced, but it does not change who is accountable for it. Practitioners still sign the work, and they still own it. The IRS is not discouraging the use of AI, but it is making clear that its use will be judged under the same standards that have always applied—just in a faster-moving and potentially more risk-prone environment.

For more information on this topic, please contact Will Covino at william.covino@fmglaw.com, Nancy Reimer at nancy.reimer@fmglaw.com, or your local FMG relationship partner.

Information conveyed herein should not be construed as legal advice or represent any specific or binding policy or procedure of any organization. Information provided is for educational purposes only. These materials are written in a general format and not intended to be advice applicable to any specific circumstance. Legal opinions may vary when based on subtle factual distinctions. All rights reserved. No part of this presentation may be reproduced, published or posted without the written permission of Freeman Mathis & Gary, LLP.