How tax professionals can harness the power of AI while maintaining professional judgment

AI is no longer a distant technological frontier — it’s already in your office. From research platforms to document review tools and advanced analytics software, AI is woven into how modern tax and accounting firms operate. Most practitioners use AI-powered tools daily without thinking twice about it.

But today’s AI tools, particularly generative AI systems that create original content, present a different challenge. In this context, “AI” refers primarily to modern AI technologies used in tax practice, including generative AI tools. But some of the risks and regulatory considerations discussed apply specifically to generative AI and in that case are identified as such.

As AI becomes more common in tax practice, questions about professional responsibility, accuracy, confidentiality, and oversight have moved to the forefront. And, the IRS has taken notice — in June 2026, the Office of Professional Responsibility issued Bulletin 2026-19, “Introductory Guidelines for Responsible AI Use in Federal Tax Practice.” This isn’t guidance for the future; it’s the regulatory framework for now.

Unlike traditional AI that assists with defined tasks, generative AI operates with discretionary decision-making capability. And that distinction matters enormously for tax professionals. Understanding what AI can and cannot do isn’t optional anymore; it’s essential to your practice and explicitly required by the IRS.

Why does this matter to you? Bulletin 2026-19 makes clear that the IRS expects competence with AI technology and holds practitioners accountable for how it’s used. More importantly, your clients’ trust and your firm’s reputation depend on navigating this new landscape thoughtfully.

Jump to ↓

The IRS has spoken: Bulletin 2026-19 changes everything

What is generative AI and how does it actually work?

What Bulletin 2026-19 means for tax practitioners using professional judgment

Setting your firm up for success

You’ve learned what AI is — now learn how to use it responsibly

The IRS has spoken: Bulletin 2026-19 changes everything

On June 24, 2026, the IRS’ Office of Professional Responsibility released Bulletin 2026-19 to address what it calls “a rapidly evolving technology” that “has been adopted by many tax practitioners.” The bulletin acknowledges that AI is abundant in modern tax and accounting firms— from ChatGPT to Thomson Reuters™ CoCounsel to Blue J. But it also warns that certain AI systems, particularly generative AI, introduce distinct risks that practitioners must understand and manage.

This bulletin matters because it establishes the regulatory baseline. Under Circular 230, practitioners and firms now operate under explicit expectations regarding technological competence, data security, verification procedures, and professional judgment. The IRS has documented the real-world consequences of failing to meet these obligations: court sanctions, professional discipline, reputational harm, and client notifications — and made it clear that tax practitioners are not exempt from these expectations.

The bulletin documents real-world consequences. In 2023 and 2024, courts sanctioned multiple attorneys for filing briefs containing fake citations generated by generative AI. In 2025, Deloitte Australia apologized and refunded fees after delivering a government report filled with fabricated quotes and invented references — all reportedly produced by AI.

Bulletin 2026-19 cites these cases as warnings. The message is stark: improper reliance on AI has real consequences, and the IRS expects tax practitioners to learn from these examples.

For CPAs and tax professionals, Bulletin 2026-19 is the catalyst. It signals that the regulatory environment has shifted. The question is no longer whether to use AI, but whether you’re using it in a way that meets the IRS’ standards for competence, compliance, and client protection.

What is generative AI and how does it actually work?

At its core, AI mimics human cognitive skills such as judgment, perception, and prioritization. Generative AI goes further by creating new content based on patterns learned from enormous quantities of data.

When you ask a generative AI system to draft a tax memo or summarize case law, it’s not retrieving pre-written answers from a database. It’s predicting the next word, then the next, based on patterns it learned during training.

This capability is powerful and productive. It’s also why things can go wrong. Some common mistakes include:

• Hallucinations: fabricated citations, invented case names, or false statistics presented with confidence
• Bias embedded in training data that can skew analysis or advice
• Lack of transparency about how it reached its conclusion
• Data security risks when sensitive client information is uploaded to unsecured systems
• Confidentiality violations when data from one client inadvertently influences responses about another

Understanding these limitations is critical because professional responsibility does not shift from practitioner to technology.

What Bulletin 2026-19 means for tax practitioners using professional judgment

The regulatory expectation is no longer theoretical. Bulletin 2026-19 makes it concrete. The IRS’ June 2026 guidance spells out what competence, compliance, and professional judgment look like in the AI era. It references specific Circular 230 provisions (§ 10.35, § 10.36, § 10.37) and applies them directly to AI use.

For tax practitioners, Bulletin 2026-19 creates three new imperatives:

1. Understand what AI can and cannot reliably do.
2. Establish firmwide procedures demonstrating compliance with Circular 230.
3. Maintain documented verification processes showing that professional judgment remains firmly in human hands.

By applying existing Circular 230 requirements to AI-enabled workflows, the bulletin reinforces that firms remain responsible for accuracy, oversight, professional judgment, and client protection.

Setting your firm up for success

The good news is that AI can be a legitimate competitive advantage when used properly.

Firms getting this right share common characteristics:

1. They recognize AI as a tool that amplifies expertise, not a replacement for it
2. They vet and choose enterprise-grade solutions with built-in security.
3. They establish clear protocols for staff.
4. They treat AI-generated content as a starting point, not a finished product.

Practitioners thriving in this moment aren’t avoiding AI but approaching it with both confidence and caution. They maintain rigorous verification processes and document their AI usage to demonstrate compliance.

You’ve learned what AI is — now learn how to use it responsibly

Understanding AI is only the first step. The real challenge is implementing it in a way that meets IRS expectations, protects client data, and preserves the professional judgment that defines exceptional tax practitioners.

Bulletin 2026-19 doesn’t ask tax professionals to avoid AI. It expects them to demonstrate professional judgement, oversight, and accountability.

AI is changing what’s possible, embedding intelligence directly into your firm’s workflow so your team can focus on what matters most: judgment, analysis, and client value. Read our white paper on The power of agentic AI for modern tax and audit firms to learn more.

The firms that succeed will not be the ones that simply adopt AI. They will be the ones that adopt it responsibly.