Navigating the Algorithmic Ledger: AICPA and Former IRS Chief Danny Werfel Launch the Council on AI Risk in Tax (CART)

By Financial and Tax Policy Desk
Published in Partnership with Professional Journalistic Standards

Main Facts

The American Institute of Certified Public Accountants (AICPA), in a high-profile collaboration with former Internal Revenue Service (IRS) Commissioner Danny Werfel, has officially established the Council on AI Risk in Tax (CART). This multidisciplinary initiative brings together prominent stakeholders from across the accounting, legal, technological, governmental, and academic sectors to address the rapid, transformative integration of artificial intelligence (AI) within tax administration and practice.

As artificial intelligence shifts from a theoretical luxury to a core operational strategy across financial services, the tax sector faces unprecedented opportunities alongside severe compliance, liability, and ethical challenges. CART is engineered to confront these challenges head-on. The coalition includes major regulatory and professional bodies, such as the National Association of Enrolled Agents (NAEA) and the Federation of Tax Administrators (FTA).

The primary mission of CART is to refine and operationalize existing risk frameworks, develop practical methodologies for responsible AI deployment, and ensure that information integrity, accountability, and robust risk management remain the cornerstones of modern tax practice. This initiative arrives at a critical juncture, coinciding directly with emerging regulatory tension between tax practitioners and federal authorities regarding how the economic efficiencies and cost savings generated by AI should be distributed between firms and their clientele.

Chronology

To understand the urgent impetus behind the formation of the Council on AI Risk in Tax, it is necessary to examine the timeline of technological integration and regulatory friction that has characterized the tax sector over recent years:

• Late 2022 – 2023: Generative AI platforms and large language models (LLMs) experience an exponential surge in public availability. Tax and accounting software developers rapidly integrate machine learning algorithms into data intake, audit preparation, research, and client communication tools.
• Early 2024: Tax practitioners begin widely adopting AI tools to streamline workflows, reduce administrative burdens, and optimize preparation timelines. However, concerns regarding data privacy, hallucinations in tax research, and professional liability begin to mount among major accounting bodies.
• September 2025 (Early Month): Recognizing the lack of cohesive industry standards, the AICPA publishes an initial AI risk framework designed to serve as a foundational baseline for CPAs and financial professionals utilizing machine learning in federal tax practice.
• September 2025 (Mid-Month): The IRS Office of Professional Responsibility issues formal guidelines citing Treasury Circular 230 (31 C.F.R. Part 10), suggesting that tax practitioners are obligated to pass AI-related cost savings directly on to their clients to avoid charging "unconscionable fees."
• September 2025 (Late Month): The AICPA pushes back against the IRS guidance, formally seeking clarification. The Institute argues that the IRS language fails to recognize the massive upfront capital investments, ongoing software maintenance costs, and liability risks associated with implementing compliant AI systems, while also ignoring modern value-pricing models.
• October 2025 – Present: In response to escalating regulatory uncertainty and technological disruption, the AICPA partners with former IRS Commissioner Danny Werfel to officially launch CART. The council schedules a series of rolling sessions throughout the year to build upon the September risk framework, bridge the gap between tax practice and federal oversight, and formulate standardized guidelines for the industry.

Supporting Data and Ecosystem Scope

The integration of artificial intelligence into tax compliance and advisory services is no longer a peripheral trend; it represents a fundamental restructuring of the financial sector. According to market data highlighted by industry groups, the adoption curve for automation tools in accounting has steepened dramatically.

Key Pillars of the CART Ecosystem

CART’s membership reflects a cross-section of the entire tax ecosystem, designed to ensure that no single perspective dominates the discourse:

1. Accounting and Advisory: Represented by the AICPA and independent public accounting firms ranging from regional practices to Big Four networks, focusing on operational feasibility and client service quality.
2. Enrolled Agents: Represented by the National Association of Enrolled Agents (NAEA), bringing ground-level insights from practitioners who navigate direct IRS representation and compliance daily.
3. State and Federal Administration: Represented by the Federation of Tax Administrators (FTA) and former high-ranking government officials like Danny Werfel, ensuring alignment with state revenue departments and federal expectations.
4. Legal and Academic Sectors: Legal scholars and technologists who evaluate liability, intellectual property, data privacy, and algorithmic transparency.

The Technological Imperative

The technological scope of CART addresses several specific operational areas where AI intersects with tax:

• Automated Document Processing: Natural language processing (NLP) models reading complex 1099s, K-1s, and corporate ledger statements.
• Predictive Compliance Analytics: Machine learning algorithms flagging audit triggers and anomalies prior to electronic filing.
• Tax Research Assistants: Generative models synthesizing thousands of pages of the Internal Revenue Code, Treasury regulations, and court cases in seconds.

Despite these efficiencies, the risk matrix remains dense. Hallucinations in tax code interpretation can lead to systemic under-reporting or severe penalties for taxpayers. Furthermore, the ingestion of sensitive financial Personally Identifiable Information (PII) into cloud-based AI environments creates massive cybersecurity vulnerabilities that demand rigorous governance frameworks.

Official Responses and Stakeholder Perspectives

The launch of the Council on AI Risk in Tax has drawn strong statements from key leaders across the tax community, emphasizing both the promise and the peril of the technological shift.

Danny Werfel on the Stakes of Self-Regulation

Former IRS Commissioner Danny Werfel, serving as co-leader of the initiative, underscored the high-stakes nature of the tax environment and the impossibility of relying on passive adaptation.

"Tax is high stakes, and we cannot assume that emerging AI tools will police themselves," Werfel stated during the announcement. "The AI risk framework published earlier this month provides a starting point. Through this new council, the AICPA is bringing the tax community together to refine that framework and explore additional tools that can help tax administrators and practitioners realize the benefits of AI while managing the various and material risks that AI presents."

Werfel’s involvement lends immense regulatory credibility to CART, signaling to both federal authorities and private practitioners that the initiative is committed to practical, enforceable standards rather than abstract theory.

Melanie Lauridsen on Strategic Necessity and Governance

Melanie Lauridsen, the AICPA’s Vice President of Tax Policy & Advocacy, emphasized that AI has evolved past the point of being a mere optional upgrade. For modern accounting practices, it is a strategic necessity.

"AI has the potential to transform how tax is administered and practiced but requires thoughtful governance, practical risk management, and trusted guidance," Lauridsen explained. "That’s why this council was launched, bringing together leaders from across the tax ecosystem to evolve the recently published risk framework, but also to develop additional tools and insights needed to adopt AI responsibly while preserving trust, accountability, and confidence in tax."

Lauridsen noted that while efficiency gains are real, they must be balanced against the absolute requirement to maintain public trust in the integrity of the nation’s tax system.

Implications: The Regulatory Clash Over AI Cost Savings and Circular 230

The formation of CART does not occur in a vacuum; it is directly shadowed by an ongoing policy dispute between the tax profession and the federal government regarding the economic rewards and regulatory penalties of AI adoption.

At the heart of this friction is guidance issued by the IRS Office of Professional Responsibility. The office released regulatory interpretations pointing directly to Treasury Circular 230—the federal regulations governing practice before the Internal Revenue Service, specifically 31 C.F.R. Part 10—which strictly prohibits practitioners from charging "unconscionable fees." The IRS guidance suggests that because AI tools drastically reduce the time and labor required to complete tax returns and advisory tasks, those cost savings must be systematically passed along to clients in the form of lower fees.

The AICPA Counter-Perspective

The AICPA has strongly pushed back against this interpretation, formally seeking clarification from the IRS. The Institute’s advocacy division points out several fundamental flaws in the federal government’s current stance:

• Ignorance of Upfront Capital and Maintenance Costs: Implementing secure, enterprise-grade AI infrastructure requires substantial financial investment in software licensing, data security protocols, continuous staff training, and specialized IT personnel. These expenses offset immediate operational efficiencies.
• The Risk Mitigation Premium: Unlike traditional software, generative AI models carry inherent legal and financial liabilities, including the risk of compliance errors or data breaches. Practitioners assume these risks on behalf of their clients, which warrants appropriate compensation.
• Incompatibility with Value-Pricing Models: Modern progressive accounting firms frequently bill based on the value and strategic insight delivered to the client—such as complex tax planning, risk mitigation, and advisory services—rather than strict billable hours. Forcing a linear connection between automated time-reduction and client fee reduction undermines modern business models within the profession.

How CART Aims to Resolve the Impasse

By convening diverse leaders from government, legal practice, and accounting, CART serves as an ideal neutral forum to negotiate these complex economic and regulatory questions. Over the coming months, the council’s working groups will analyze how fee structures can evolve fairly under Circular 230 without penalizing firms for investing in innovative technologies.

Furthermore, the council’s ongoing meetings will help establish a unified industry voice in ongoing dialogues with the Treasury Department and the IRS. By proactively drafting rigorous, self-imposed risk management and governance protocols, the tax profession hopes to demonstrate that external, heavy-handed regulatory mandates are unnecessary.

Conclusion and Available Resources

As artificial intelligence continues to reshape the landscape of federal tax practice, initiatives like the Council on AI Risk in Tax will be vital in charting a sustainable course. Practitioners and firms seeking immediate guidance can utilize introductory resources provided by the AICPA, including their published guidelines and comprehensive FAQs on responsible AI use in federal tax practice.

Through continuous dialogue, rigorous risk frameworks, and collaborative leadership, the tax community aims to harness the full potential of artificial intelligence while safeguarding the integrity, accountability, and public trust that form the foundation of the global financial system.