03/07/2026
Last week, the Office of Professional Responsibility (OPR) of the IRS issued guidance on how Treasury Circular 230 rules apply to tax practitioners’ use of artificial intelligence. This guidance emphasizes our duties of competence, confidentiality, and due diligence.
As artificial intelligence users in our practice at Saman Consulting, we've adopted strict workflows for reviewing every output produced by an AI model. We constantly review our Written Information Security Plan (WISP) with every model update, and all Personal Identifiable Information (PII) goes through a manual cleaning process before being used with WISP-approved AI models. We also provide our clients the choice to opt out from using their information in any AI model.
AI models should support and enhance our practice; they should never replace our professional judgment. Nothing can or should substitute our analytical capacity to interpret tax law and our ability to understand the human and economic impact of not following our professional ethical standards.
The IRS warned that AI’s risks — including fabricated outputs and data privacy concerns — require practitioners to verify results and follow existing Circular 230 rules on diligence, competence, and confidentiality.
