Policy

AICPA Challenges IRS Guidance on AI Cost Savings for Tax Clients

The accounting group says federal guidelines overlook implementation costs and the shift to value-based pricing in professional services.

Omega Editorial· September 8, 2026· 3 min read

The American Institute of CPAs is challenging recent IRS guidance that suggests tax practitioners should pass along cost savings from artificial intelligence tools directly to clients through reduced fees.

The guidance, issued in June by the IRS Office of Professional Responsibility, stated that practitioners using generative AI should reflect efficiency gains in their billing practices, specifically citing reduced research and drafting time. The office framed the recommendation within Circular 230's prohibition on unconscionable fees, describing the update as "introductory guidelines for responsible AI use in federal tax practice."

Why it matters

The dispute highlights a fundamental tension between regulators and professional services firms over how AI-driven productivity should be valued. As AI tools become standard in tax preparation and advisory work, the pricing model debate will shape both firm economics and client expectations across the accounting industry.

The AICPA's counterargument

AICPA officials argue the IRS guidance oversimplifies the economics of AI adoption. Eva Simpson, the organization's vice president for Member Value, Tax & Advisory Services, said the guidance "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."

The organization is working with the IRS on clarifying language and frequently asked questions, according to Melanie Lauridsen, AICPA's vice president for Tax Policy & Advocacy. She emphasized that value pricing has become standard business practice and must account for the liability practitioners assume.

AICPA President and CEO Mark Koziel noted that the OPR language is not authoritative and suggested it may be "overstepping a little bit." He pointed out that while some CPAs still bill by hours and rates, many have shifted to value-based pricing models.

The value-pricing shift

Jan Lewis, AICPA chair and a firm partner, stressed that AI-generated work remains valuable to clients regardless of how it's produced. "That value, however it is performed by AI or by the profession, has a cost and it has a benefit, and we need to bill for that service based on the value provided," Lewis said during a recent AICPA Town Hall.

Koziel added that firms often bundle services, including tax return preparation, at no additional charge while providing advisory services on tax, retirement, and estate planning—a direction the AICPA has actively encouraged.

Crucially, practitioners using AI still bear full risk and liability for tax return preparation and filing, according to AICPA leaders. The organization has developed its own guidelines and FAQs on AI use in federal tax practice.

The details were first reported by the Journal of Accountancy.

#artificial intelligence#tax preparation#accounting#irs#professional services pricing#aicpa

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy

Policy· 2 min read

OpenAI's Astra Model Raises Internal Safety Concerns

Chief scientist warns that frontier AI development at maximum speed poses dangers, even as new model excels at benchmarks.

Via AI Watch · Sep 8, 2026
Policy· 3 min read

Federal Court Limits Wisconsin AI Child Abuse Material Law

Appeals panel rules private possession of synthetic images may have First Amendment protection, though production and distribution remain prosecutable.

Via AI Watch · Sep 8, 2026
Policy· 3 min read

U.S. Accuses Six Chinese AI Firms of Mass IP Theft via Distillation

Federal officials allege DeepSeek, Moonshot AI, Alibaba and others systematically copied American AI models to accelerate development and cut costs.

Via AI Watch · Sep 8, 2026