The Pittsburgh Tax Review has published Symposium, AI in Tax Practice and Law School Education, 23 Pitt, Tax Rev. 283- 353 (2026) :
Sarah B. Lawsky (Illinois), Direct File as Formalization, 23 Pitt. Tax Rev. 283 (2026) (reviewed by Jeff Gordon (Vanderbilt) here):
Direct File was a program created by the United States government that allowed some taxpayers to file their federal income tax returns online, for free. This article examines the computer code that underlies Direct File and argues that Direct File is an extraordinary accomplishment that skillfully executes not a flexible formalization of the underlying statute and regulations, but rather a formalization of tax forms and informal guidance, designed to accomplish a particular task: allowing online return preparation. Direct File together with the released computer code also increases transparency beyond what forms, worksheets, and e-filing have previously allowed. The computer code reveals behind-the-scenes choices that are merely implicit in forms, instructions, and other informal guidance. The Direct File code also allows visualization and explanatory tools to be built on top of that code, making the application of the law and various administrative choices more transparent even to those who are not comfortable reading computer code. The Direct File repository is a rich source for legal research; this article only begins to scratch the surface of what can be learned from this computer code.
Kathleen DeLaney Thomas (North Carolina), Tax in the Echo Chamber, 23 Pitt. Tax Rev. 307 (2026):
The idea of an internet echo chamber — a metaphor first coined by Cass Sunstein — describes our tendency to surround ourselves online with people and content that reflect only our own viewpoints. Echo chambers not only promote polarization and extremism, but may threaten deliberation, problem-solving, and democracy. This Essay focuses on a particular type of echo chamber (sometimes called a “filter bubble”) unique to social media platforms like TikTok, Instagram, Facebook, and YouTube. One feature these platforms have in common is that they use machine-learning algorithms to select content for their users. The algorithm displays content based on what it determines will keep the user engaged on the platform the longest, which, in turn, generates the most revenue for the platform. As a result, users see content that reflects what they tend to like, including the types of videos they have watched in the past, as well as content that is generally most engaging. Content might be deemed engaging by an algorithm because it is extreme or provocative, not necessarily because it is accurate or relevant. As a result of the operation of these machine-learning algorithms, social media users often find themselves in an echo chamber of content that not only reinforces their own beliefs, but may shift them towards more extreme beliefs.
This Essay explores the concept of social media echo chambers in the context of the tax law. Specifically, the Essay argues that a significant and growing number of taxpayers encounter information about the tax system through their interaction with social media content that is selected by machine-learning algorithms. Tax information, commentary, and advice abounds on social media, and nearly 80 percent of those considered Gen-Z or Millennials seek out tax and financial advice on social media platforms. Policymakers who are interested in keeping the public informed about the tax system, thereby promoting tax compliance, must grapple with the fact that these algorithms shape how people learn tax information. As the Essay will describe, baseline tax information among U.S. taxpayers is already very low, and the potential for tax echo chambers to circulate and reinforce bad information only exacerbates this issue. What’s more, outreach efforts to educate the public are unlikely to reach users who find themselves in certain tax echo chambers.
Emily Wielk (Bipartisan Policy Center) & Chintal Shah (Bipartisan Policy Center), Automation Without Abdication: Preserving Human Oversight in AI-Driven Tax Administration, 23 Pitt. Tax Rev. 335 (2026)
