LO
ConstraintCognitive automation2026-06-24
The IRS Office of Professional Responsibility says tax practitioners must thoroughly review AI-created documents before they reach clients or the IRS, and cannot rely solely on AI
Tax preparer / tax agentoccupation page →
Event date / reported
2026-06-24
Evidence stage
ConstraintFailure, rollback, regulation or cost is suppressing adoption. Can lower an assessment or widen its uncertainty.
Tasks this bears on
Handling complicated returns
Returns for the self-employed, landlords, people with income from several countries, and small businesses.
Being augmented✓ Evidence-backed
Advising on reliefs and planning
Telling clients which reliefs they can claim, and how decisions such as selling property or starting a business will affect their tax.
Still human-led✓ Evidence-backed
Dealing with the tax authority
Answering the authority's queries, objecting to assessments and representing clients in audits.
Still human-led✓ Evidence-backed
Where this applies
The US tax authority's regulator of tax practitioners, reading existing Circular 230 duties for generative AI. It says that when using generative AI, practitioners must thoroughly review all AI-created documents and language before delivery to a client or submission to the IRS, verifying facts, citations and calculations; that practitioners cannot rely solely on AI; and that when drafting written advice they must independently authenticate all factual and legal information. It applies existing duties rather than creating a new rule, it governs how AI is used rather than whether, and it gives no figures on how many practitioners use AI.
What this means
The regulator of US tax practitioners has put in writing that AI output does not carry the practitioner's responsibility: every document and every piece of advice drafted with AI has to be checked by the person who signs it. Advice, complex returns and dealings with the IRS stay the practitioner's to answer for.
What it does not yet show
It restates existing duties for one jurisdiction; it does not measure how many practitioners use AI or how often AI-drafted work is wrong.
What you can check
Open IRS Office of Professional Responsibility Issue Number 2026-19 and find "Practitioners cannot rely solely on AI".
Does it change the assessment?
No. The impact index is never moved by a single event. What this record did: the 3 linked task judgements above now rest on evidence instead of inference.
Source
Internal Revenue Service, Office of Professional Responsibility — Issue Number 2026-19, Introductory Guidelines for Responsible AI Use in Federal Tax Practice (June 24, 2026) · verified 2026-09-27 · Claude (VOLO agent) · interpreted 2026-09-27 · Claude (VOLO agent)
Primary source — published by the party that did this, or the authority of record. No co-signature needed.