IRS Issues First AI Guidelines for Tax Practitioners Under Circular 230

The IRS Office of Professional Responsibility has issued its first comprehensive guidance on artificial intelligence use in federal tax practice, establishing clear standards for practitioners deploying generative AI tools while maintaining compliance with Treasury Department Circular 230.

The alert, published June 24, 2026, addresses the fundamental shift from traditional analytical software to generative AI systems that can make autonomous decisions and learn from their own outputs. According to the OPR, these platforms use "complex pattern-recognition capabilities that continually interact and evolve," creating both efficiency opportunities and significant regulatory risks.

Why it matters

Tax professionals face mounting pressure to adopt AI for competitive advantage, but the technology's technical flaws—fabricated citations, algorithmic bias, and data cross-contamination—have already triggered court sanctions, financial penalties, and disciplinary referrals. This guidance establishes that efficiency gains don't reduce professional liability; practitioners remain fully accountable for all AI-generated work product submitted to clients or the IRS.

Core compliance obligations remain unchanged

The OPR emphasizes that AI deployment doesn't alter foundational duties under Circular 230. Section 10.22's due diligence requirement mandates that practitioners "thoroughly review all AI-created documents and language" before delivery, including verification of facts, citations, and calculations. Human scrutiny cannot be delegated to algorithms.

Section 10.35's competence standard now explicitly encompasses technological literacy. Practitioners must understand "AI systems' operational mechanics, limitations, and risks," including how platforms develop content and the potential for bias or errors. The guidance makes clear that legal knowledge alone is insufficient—professionals must evaluate whether AI outputs are suitable for IRS matters.

Firm-level controls and systemic risk

Firm leaders face affirmative obligations under Section 10.36 to implement adequate AI compliance procedures. Managing practitioners must establish staff training programs addressing GAI risks, protocols for secure data handling, and vetting processes for third-party AI tools. Failure to maintain these controls through "willfulness, recklessness, or gross incompetence" creates direct disciplinary exposure.

The guidance addresses practical billing concerns under Section 10.27, warning that charging clients for manual labor time not actually spent—or double-billing for AI-assisted tasks—may constitute unconscionable fees. Cost savings from AI efficiency should be transparently reflected in billing practices.

Data security and statutory penalties

Beyond Circular 230, practitioners face civil and criminal liability under IRC Sections 6713 and 7216 for unauthorized disclosure of tax return information. The OPR warns that uploading client data to unsecured or public AI systems creates acute exposure. Only enterprise-approved platforms with robust confidentiality safeguards satisfy federal requirements.

The guidance notes that courts have already imposed financial sanctions, public censure, mandatory ethics training, and bar referrals for AI-related misconduct. In one 2025 case, Deloitte Australia was forced to partially refund fees after delivering a government report containing AI-generated fabricated judicial quotes and nonexistent source citations.

Evolving regulatory landscape

The federal guidance arrives as multiple states—including California, Colorado, Illinois, and Utah—enact AI governance legislation focused on transparency and consumer protection. The American Bar Association's July 2024 Formal Opinion 512 on generative AI tools similarly reinforces that technological advancement doesn't diminish core professional duties.

The IRS guidance was first reported by Current Federal Tax Developments and emphasizes that while AI offers transformative potential for tax practice efficiency, it remains "a powerful tool, not a substitute for professional judgment." Practitioners must maintain ongoing education as regulatory frameworks continue to evolve.