Blab AI
IRS Faces AI Management Gaps After GAO Report Highlights Staffing and Inventory Shortfalls
The Internal Revenue Service (IRS) is expanding its use of artificial intelligence (AI) to flag tax returns for additional review, but a recent Government Accountability Office (GAO) report shows the agency is not yet prepared to scale the technology.
The GAO’s March 2026 report, GAO‑26‑107522, examined the IRS’s AI programs and found that the agency lacks the skilled workforce and a complete inventory of the AI systems it currently operates. The report notes that the IRS has 126 active AI use cases—specific applications of AI for audit selection, fraud detection, and taxpayer support—as of June 2025, but it could not produce a full inventory of those systems.
AI use cases are defined by the IRS as scenarios in which AI is designed, developed, procured, or used to advance the agency’s mission. The technology is intended to help the IRS identify discrepancies in returns and research precedents faster than human staff can. When an AI system flags a return, a human IRS employee reviews the case.
The GAO report identified eight recommendations for the IRS commissioner. They include: identifying the skills the agency needs and developing a plan to address gaps; creating a quality‑assurance process for AI inventory entries; developing comprehensive internal AI guidance; clarifying communications about unclassified AI use cases; identifying all existing AI contracts and notifying responsible parties of governance requirements; implementing policies to increase internal coordination and avoid duplication; requiring AI case owners to report how use cases align with strategic goals; and establishing performance metrics for AI outcomes.
The IRS has agreed to all eight recommendations and will provide status updates in the fall. The agency’s acceptance follows a broader Treasury Department effort to standardize AI governance. Treasury’s Artificial Intelligence System User Agreement, published October 2025, requires IRS AI use to conform to federal standards.
The GAO’s findings are significant because the IRS has faced budget cuts and workforce reductions in recent years. The agency’s staff shortages have limited its ability to develop and maintain AI tools. The GAO report warns that without addressing these gaps, the IRS could experience increased errors in audit selection or taxpayer service.
The IRS’s use of AI is part of a broader effort to modernize tax administration. According to the IRS’s own documentation, AI is used for audit selection, fraud detection, and answering taxpayer questions through voicebots and chat interfaces. The agency’s AI policy, IRM 10.24.1, outlines the governance framework for AI use.
While the GAO report does not detail specific performance metrics for the IRS’s AI systems, it emphasizes the need for a clear inventory and quality‑assurance process. The report also notes that several IRS divisions use different AI systems, creating potential overlap and duplication.
The GAO’s assessment comes at a time when the IRS is under pressure to improve compliance and revenue collection while operating with a leaner workforce. The agency’s plan to address the identified gaps will be closely watched by Congress and taxpayers.
In summary, the IRS is expanding AI use to improve audit selection and taxpayer service, but a GAO report highlights critical deficiencies in staffing, inventory, and governance. The agency’s agreement to the GAO’s recommendations and its planned fall updates will determine whether it can effectively manage its AI programs without compromising audit quality or taxpayer experience.
The IRS’s next steps include developing a workforce plan to fill AI skill gaps, establishing a formal inventory and quality‑assurance process, and aligning AI use cases with strategic goals. The outcome of these efforts will shape the agency’s ability to leverage AI for tax administration in the coming years.
The GAO’s March 2026 report, GAO‑26‑107522, examined the IRS’s AI programs and found that the agency lacks the skilled workforce and a complete inventory of the AI systems it currently operates. The report notes that the IRS has 126 active AI use cases—specific applications of AI for audit selection, fraud detection, and taxpayer support—as of June 2025, but it could not produce a full inventory of those systems.
AI use cases are defined by the IRS as scenarios in which AI is designed, developed, procured, or used to advance the agency’s mission. The technology is intended to help the IRS identify discrepancies in returns and research precedents faster than human staff can. When an AI system flags a return, a human IRS employee reviews the case.
The GAO report identified eight recommendations for the IRS commissioner. They include: identifying the skills the agency needs and developing a plan to address gaps; creating a quality‑assurance process for AI inventory entries; developing comprehensive internal AI guidance; clarifying communications about unclassified AI use cases; identifying all existing AI contracts and notifying responsible parties of governance requirements; implementing policies to increase internal coordination and avoid duplication; requiring AI case owners to report how use cases align with strategic goals; and establishing performance metrics for AI outcomes.
The IRS has agreed to all eight recommendations and will provide status updates in the fall. The agency’s acceptance follows a broader Treasury Department effort to standardize AI governance. Treasury’s Artificial Intelligence System User Agreement, published October 2025, requires IRS AI use to conform to federal standards.
The GAO’s findings are significant because the IRS has faced budget cuts and workforce reductions in recent years. The agency’s staff shortages have limited its ability to develop and maintain AI tools. The GAO report warns that without addressing these gaps, the IRS could experience increased errors in audit selection or taxpayer service.
The IRS’s use of AI is part of a broader effort to modernize tax administration. According to the IRS’s own documentation, AI is used for audit selection, fraud detection, and answering taxpayer questions through voicebots and chat interfaces. The agency’s AI policy, IRM 10.24.1, outlines the governance framework for AI use.
While the GAO report does not detail specific performance metrics for the IRS’s AI systems, it emphasizes the need for a clear inventory and quality‑assurance process. The report also notes that several IRS divisions use different AI systems, creating potential overlap and duplication.
The GAO’s assessment comes at a time when the IRS is under pressure to improve compliance and revenue collection while operating with a leaner workforce. The agency’s plan to address the identified gaps will be closely watched by Congress and taxpayers.
In summary, the IRS is expanding AI use to improve audit selection and taxpayer service, but a GAO report highlights critical deficiencies in staffing, inventory, and governance. The agency’s agreement to the GAO’s recommendations and its planned fall updates will determine whether it can effectively manage its AI programs without compromising audit quality or taxpayer experience.
The IRS’s next steps include developing a workforce plan to fill AI skill gaps, establishing a formal inventory and quality‑assurance process, and aligning AI use cases with strategic goals. The outcome of these efforts will shape the agency’s ability to leverage AI for tax administration in the coming years.
