Your AI Doesn't Have a PTIN: Due Diligence Under Circular 230

About the course

Circular 230 obligations attach to the practitioner regardless of what tool produces the work. When AI drafts a memo, summarizes a Regulation, codes a Schedule C, or answers a client question, §10.22 due diligence, §10.35 competence, §10.37 written advice standards, §7216 confidentiality, §10.34 return positions, §6694 preparer penalties, and the §10.36 firm compliance duty all apply to the output.

This course covers the practitioner obligations that apply to AI-assisted tax work, using a running fact pattern of an enrolled agent integrating AI tools into a small practice. Coverage includes where each obligation attaches, what a defensible file contains, and where the exposure sits when a return goes out that the tool got wrong. The course closes with the five minimum written controls a practice needs before AI touches a client file.

Two hours of Ethics credit under Circular 230.

Learning objectives

By the end of this course, participants will be able to:

1. Identify the Circular 230 provisions that govern practitioner use of AI tools in tax practice.
2. Distinguish between the tool's output and the practitioner's use of the output for §10.22 due diligence purposes.
3. Recognize the §10.34 return position standards and §6694 preparer penalty framework as they apply to AI-generated positions.
4. Recognize the §7216 and §6713 confidentiality obligations triggered when client data is submitted to third-party AI systems.
5. Identify the minimum written controls required under §10.36 before integrating AI into a tax workflow.

Syllabus

1. SECTION 01What the tool cannot hold

   Why Circular 230 attaches to the practitioner, not the technology, and how the licensing structure means the tool's speed doesn't reduce the practitioner's obligation.

2. SECTION 02Due diligence under §10.22

   What "reasonable reliance" means when the source is an AI output, and what verification the record must show.

3. SECTION 03Competence under §10.35

   The distinction between using a tool competently and having competence in the underlying subject matter.

4. SECTION 04Written advice under §10.37

   How the §10.37 standards apply when the practitioner adopts an AI-drafted memo, and what the file needs to show.

5. SECTION 05Confidentiality under §7216 and §6713

   When submitting client data to an AI service constitutes disclosure, and what consent — if any — protects the practitioner.

6. SECTION 06Return positions and preparer penalties

   §10.34 return standards, §6694 preparer penalties, and how AI-generated positions get evaluated after the fact.

7. SECTION 07The five written controls

   The minimum controls §10.36 requires in a practice using AI tools, and what defensible documentation looks like.

Instructor

Jennifer Harris Smith, JD, CPA. Attorney (Texas). CPA (Texas). Member of Texas Bar College. Full bio →