The Journal of Accountancy has published an interesting article about comments made by the director of the IRS Office of Professional Responsibility at the AICPA National Tax Conference yesterday, an article that you should probably read if you’re in practice and using AI tools.
First, let’s get this one out of the way since it’s a big topic and one the AICPA has specifically spoken out about:
Thomas Curtin, who noted that he was speaking in his personal capacity and not on behalf of the IRS, said the OPR was “not suggesting that practitioners must reduce fees simply because AI allows you to complete certain tasks more effectively.”
If you missed the earlier discussion, OPR issued its first loose AI guidance back in June and the AICPA got fairly bent out of shape over this section:
AI can reduce research and drafting time, such that billing clients for manual labor or time that was not actually spent or double billing for AI-assisted tasks may violate § 10.27, depending on the facts (e.g., a noticeable pattern across clients or the size of the billing differentials). Cost savings should be passed on openly, with billing practices that reflect the efficiencies gained from the use of GAI. Practitioners should not only disclose, in general or specific terms as needed, the AI activities performed, but also fairly credit to the client’s account any cost reductions.
“The IRS statement implies that any efficiencies gained through AI should be directly passed on to the client,” said Eva Simpson, CPA, CGMA, AICPA vice president–Member Value, Tax & Advisory Services, in September. “That is an overly simplistic view that ignores the full economics of AI adoption, including software licensing costs, implementation expenses, governance requirements, and the significant investment needed to train professionals to use these tools responsibly and effectively. It also overlooks that professional services are increasingly priced based on value delivered, not just time spent, and that the benefits AI creates for clients should be considered as part of that equation.”
On Monday, Curtin expressed that he felt it was “worthwhile to kind of get the conversation going” and added: “And again, they’re just guidelines. I’m not up here trying to say that we’re creating new authority or anything like that.”
A few other things he and panelist Ryan Corcoran, who is a partner at RSM, said during the session, as reported by JofA:
- The two warned that taxpayers and practitioners should be cautious about entering tax return information into public AI tools because of potential confidentiality and Sec. 7216 concerns.
- Curtin said AI is not considered a “person” under Circular 230, meaning practitioners cannot rely on AI-generated work in the same way they may rely on the work of another professional.
- Corcoran highlighted concerns about “shadow AI,” in which employees use unapproved AI tools outside firm-approved systems and security controls.
- Corcoran said that many practitioners now receive AI-generated tax analyses from clients and must review those materials using the same professional judgment they apply to other client-provided information.
OPR has heard the complaints and more guidance shall come in the months ahead. Surely they’ll be tiptoeing around any discussions around fees as they relate to AI efficiencies and time savings.
