Jay A. Soled (Rutgers; Google Scholar) & Kathleen DeLaney Thomas (North Carolina; Google Scholar), Predictive Analytics and the Tax Code, 51 Fla. St. U. L. Rev. 597 (2024) (reviewed by Michelle Layser (San Diego; Google Scholar) here):

Congress last reformed the nation’s current tax penalty regime approximately three decades ago, long before the rise of big data and the advent of predictive analytics. With predictive analytics now gaining preeminence and its accuracy constantly improving, it is time for Congress to weave this technological innovation into the fabric of the Internal Revenue Code and, more specifically, the civil tax penalty regime. Doing so would enhance taxpayer compliance, augment transparency, and simultaneously ease many administrative burdens commonplace under the tax law.

Conclusion
A central feature of every tax system is a viable penalty regime. Absent one, many taxpayers might not fulfill their civic duties and, as a result, revenue collections would likely diminish. On the other hand, the better and more efficient a tax system’s penalty regime, the greater the odds that the particular tax in question will flourish and result in the generation of sufficient funds to meet public expenditures.

For over a century, the nation’s income tax has relied upon the existing penalty system to keep taxpayers’ tax reporting positions in check, punctuated by periodic reform measures—the most recent being over three decades ago. And, at least by numerical standards, the existing penalty regime has done an admirable job. With little fanfare, the IRS routinely collects trillions of dollars annually. Furthermore, the nation’s voluntary compliance rate is one of the highest in the industrialized western world.

But the advent of AI coupled with predictive algorithms opens the door to vast reform opportunities. When auditing taxpayers’ tax returns, the IRS can employ AI and predictive algorithms to determine, in light of the proffered probability thresholds and when it comes to subjective determinations, whether penalty imposition is apropos. However, Congress should consider instituting a safe harbor protection for circumspect taxpayers who utilize twenty-first-century technology to ascertain whether their reporting positions meet or exceed governing probability thresholds. Instituting these measures that incorporate AI and predictive algorithms into the fold of the tax system will yield far greater tax compliance and, in addition, transparency.

Technological advancements will admittedly not solve all of the tax collection issues besetting the nation’s tax system. Insofar as tax penalties are concerned, however, AI and predictive analytics open a new chapter in tax administration, offering a powerful compliance platform that, aside from being extraordinarily quick and cost-efficient, is transformative in nature. Their adoption would constitute a major step forward in tax administration, one that all parties—namely, Congress, the IRS, taxpayers, and tax practitioners—would be wise to quickly embrace. 

Editor's Note:  If you would like to receive a daily email with links to tax posts on TaxProf Blog, email me here.