Gregory T Bryant, CPA and Attorney at Law
Caroline White, JD
DISCLAIMER: This article is for informational purposes only and does not constitute legal, tax, or professional advice. Readers should consult with qualified professionals regarding their specific circumstances.
Introduction
In this article we are looking into our crystal ball and asking, “Where does the Internal Revenue Service go from here?”. The Internal Revenue Service faces significant operational challenges, including staffing constraints, personnel shortages, inefficiencies, and aging technological infrastructure. As of the date of this article, the IRS operates with aging software systems (some dating back approximately 40 years) and has experienced significant staffing reductions, many of whom are close to retirement. Where does the IRS go from here?[1]
We believe a significant path forward for the IRS involves adopting and deploying advanced technology based on artificial intelligence (“AI”). AI software solutions to collect data, process and match W-2s, 1099s, K-1s, 1098s to income tax returns, provide taxpayer services and problem resolution, select audit candidates, track down fraud, enforce liens and garnishments and more. For taxpayers, this might mean paperless tax filings, less opportunity to cheat, and more automated enforcement, from a tireless, relentless machine that works 24 x 7.
The Internal Revenue Service is entering a new era of tax enforcement and customer service powered by artificial intelligence and as a result of DOGE and shutdown, that process needs to accelerate. While some see this as a much-needed step toward efficiency and fairness in the tax system, others worry about privacy, accuracy, and the risks of relying too heavily on algorithms. In this article, we’ll explore how the IRS is using AI today, where it’s headed, and what it means for taxpayers and businesses navigating an increasingly digital tax landscape.
The Current State
The IRS currently uses AI primarily for audit selection and collection efforts targeting high-income taxpayers and complex partnerships. As of February 2024, according to publicly available reports, the agency was running sixty-eight AI-related initiatives: twelve focused on customer service, twenty-seven on enforcement, and twenty-nine on overall operations. Looking ahead to 2025, one of the IRS’s stated priorities is to “develop and implement an Artificial Intelligence capability” for handling referrals concerning tax-exempt entities. The Treasury Department has recently announced new tech contracts that should further extend the use of AI at the IRS in the future.
Despite these efforts, the IRS has faced criticism regarding the accuracy and effectiveness of its algorithms, as well as concerns about the necessary data access required to maximize AI use. In 2024, the Government Accountability Office (GAO) highlighted shortcomings in a study on the IRS’s AI initiatives, noting the agency’s failure to “document key components and technical specifications” of its pilot AI method for 1040 audit selection (GAO-24-106449). Furthermore, the IRS remains cautious about sharing AI-related information, citing worries that taxpayers could exploit such disclosures to circumvent enforcement efforts.
Looking to the Future
Following recent workforce reductions from the Department of Government Efficiency (DOGE) and now “Shutdown”, the IRS is virtually rebuilding from the ground up. It is the perfect time to adopt new, innovative solutions. The need to integrate artificial intelligence into core operations has become increasingly urgent. With fewer personnel to handle a growing volume of taxpayer inquiries and administrative tasks, AI offers a potential pathway to maintain service levels and achieve greater efficiency; However, workforce reductions have also resulted in the loss of experienced personnel with institutional knowledge of the current system’s operational challenges.
The use of AI is not only an operational necessity for the future, but a strategic one as well. AI technology can be used for both legitimate and malicious purposes, and there are concerns that it could be used to identify vulnerabilities in current IRS systems, making it important that the IRS continues to safeguard sensitive taxpayer data against advanced threats. Additionally, recent discussions around the Treasury’s and IRS’s Criminal Investigation division’s expanded capabilities highlight how AI tools could extend the agency’s analytical capabilities beyond traditional tax return analysis.
However, the IRS’s AI implementation faces challenges in fully realizing this potential. The implementation process has faced criticism for lacking coordination, despite gradual progress. Persistent concerns over data privacy, bias, and accuracy remain, in part because the IRS has disclosed limited details about its AI operations. Greater transparency could help address these concerns, though the agency has cited legitimate concerns about taxpayers exploiting such disclosures to circumvent enforcement efforts. Additionally, legacy systems that date back decades were not designed to support modern technology; however, the Treasury Department also announced a contract with Palantir to overhaul the current system so that it can function in the current technological environment.
AI holds clear potential to transform IRS operations, particularly by improving efficiency and customer relations. With reduced staffing levels and increased operational pressures, there may be greater organizational openness to technological innovation. While current computer systems already process returns, they often struggle to identify even minor anomalies and if a payment is misapplied to a period or tax identification number, it is extremely hard to fix it. AI’s learning capabilities could enhance these systems, reduce human error, and lessen the workload for agents who must correct system failures manually. The Treasury Department has entered into four new contracts to further advance the “Zero Paper Initiative” and move even more taxpayer interaction online. A well-designed AI system could potentially analyze audit-selection factors more comprehensively than traditional methods, potentially reducing reliance on randomized sampling by evaluating returns for audit characteristics more systematically.[2]
Artificial Intelligence cannot be adopted and deployed without a solid foundation under it. This means people with institutional knowledge and experience must actively participate in the adoption of AI. AI also needs good data. In applications of AI, legacy data may not be adequate to do proper analytics. The solution is to update, and upgrade legacy data. Also, technology solutions cannot solve personnel problems. Personnel processes and automation work together. You do not want to build highways over goat paths.
Conclusion
The IRS’s adoption of artificial intelligence represents both a significant opportunity and a complex challenge. While AI is necessary to enhance efficiency, improve audit accuracy, and streamline operations, its current implementation highlights the need for careful oversight, transparency, and robust safeguards. Moving forward, the agency’s ability to balance technological innovation with privacy, fairness, and accountability will be crucial in shaping a tax system that benefits both the government and taxpayers. Recent government partnerships with leading technology firms represent a significant step toward accelerating modernization efforts while managing the inherent risks involved.
[1] DISCLAIMER: The statements in this article represent the author’s opinions and predictions and should not be construed as statements of fact or guarantees of future outcomes.
[2] NOTE: Any AI-based audit selection system would need to comply with applicable laws and regulations, including protections against discriminatory practices and requirements for due process.
Sources
https://www.jdavidtaxlaw.com/blog/how-the-irs-is-utilizing-ai-to-pursue-tax-debts-in-2025/
https://www.gao.gov/products/gao-24-106449
https://www.gao.gov/assets/gao-24-106449.pdf
https://www.taxnotes.com/featured-news/transparency-oversight-urged-irs-artificial-intelligence/2024/10/21/7m6nv
https://tax.thomsonreuters.com/news/amid-budget-cuts-ai-could-play-role-in-irs-modernization/
https://home.treasury.gov/news/press-releases/sb0260
https://www.foxnews.com/politics/riot-inc-trump-launches-whole-of-government-push-expose-antifa-funding-networks-dark-money-sources
https://home.treasury.gov/news/press-releases/sb0260