Andrew Blair-Stanek’s research and teaching interest is tax law. He draws from non-tax scholarship to improve tax law. His focus is developing artificial intelligence tools to extract computer-usable meaning from the tax code, treasury regulations, IRS rulings, and tax case law – with the ultimate goal of allowing the IRS to automatically identify possible tax avoidance.

Prior to joining the University of Maryland Francis King Carey School of Law faculty, Professor Blair-Stanek practiced tax law at McDermott, Will & Emery, LLP in Washington, DC. His practice included bankruptcy taxation, intellectual property transactions, cross-border mergers, international tax planning, and real estate investment trust formation. Prior to practice, he clerked in Baltimore for the Hon. Paul V. Niemeyer, U.S. Court of Appeals for the Fourth Circuit.

Professor Blair-Stanek received his undergraduate degree in mathematics, summa cum laude, from Princeton University, and his JD from Yale Law School, where he was on the Yale Law Journal. Before attending law school, he worked as a software design engineer for Microsoft Corporation and is the inventor of U.S. Patents 7,617,204 and 7,580,951. He is currently pursuing a PhD in computer science at Johns Hopkins University.

Articles

LLMs Provide Unstable Answers to Legal Questions, ICAIL '25: Proceedings of the Twentieth International Conference on Artifical Intelligence and Law (forthcoming) (with Benjamin Van Durme). Abstract

BLT: Can Large Language Models Handle Basic Legal Text?, Proceedings of the Natural Language Processing Workshop 216 (2024) (with Nils Hozenberger and Benjamin Van Durme). Abstract

Can GPT-3 Perform Statutory Reasoning?, Proceedings of the 19th International Conference on Artificial Intelligence & Law (2023) (with Nils Holzenberger & Benjamin Van Durme), https://dl.acm.org/doi/10.1145/3594536.3595163. Abstract

OpenAI Cribbed Our Tax Example, But Can GPT-4 Really Do Tax?, 180 Tax Notes Federal, Aug. 14, 2023, at 1101 (with Nils Holzenberger & Benjamin Van Durme). Abstract

Improved Induction of Narrative Chains via Cross-Document Relations, Proceedings of the 11th Joint Conference on Lexical and Computational Semantics 208 (2022) (with Benjamin Van Durme) Abstract

Shelter Check: Proactively Finding Tax Minimization Strategies via AI, 177 Tax Notes Federal, Dec. 12, 2022, at 1515 (with Nils Holzenberger & Benjamin Van Durme). Abstract

AI for Tax Analogies and Code Renumbering, 170 Tax Notes Federal, Mar. 29, 2021, at 1997 (with Benjamin Van Durme). Demo available at http://taxanalogies.law.umaryland.edu/ Abstract

Contractual Tax Reform, 61 William & Mary Law Review 1537 (2020) (with Michael Abramowicz). Abstract

How the IRS Should Fight the COVID-19 Economic Crisis, Tax Notes Federal, Mar. 30, 2020, at 2067. Abstract

A Dataset for Statutory Reasoning in Tax Law Entailment and Question Answering, Proceedings of the 2020 Natural Legal Language Processing (NLLP) Workshop (2020) (with Nils Holzenberger and Benjamin Van Durme). Abstract

Crises and Tax, 67 Duke Law Journal 1155 (2018). Abstract

Explaining the Enigmatic Expulsion: Northwest Wholesale v. Pacific Stationery, 53 Willamette Law Review 335 (2017). Abstract

Just Compensation as Transfer Prices, 58 Arizona Law Review 1077 (2016). Abstract

Intellectual Property Law Solutions to Tax Avoidance, 62 UCLA Law Review 2 (2015). Abstract

Tax in the Cathedral: Property Rules, Liability Rules, and Tax, 99 Virginia Law Review 1169 (2013). Abstract

Twombly is the Logical Extension of the Mathews v. Eldridge Test to Discovery, 62 Florida Law Review 1 (2010). Abstract

Increased Market Power as a New Secondary Consideration in Patent Law, 58 American University Law Review 707 (2009). Abstract

Note, Profits as Commercial Success, 117 Yale Law Journal 642 (2008). Abstract

Note, Using Insurance Law and Policy to Interpret the Tax Code’s Loss and Medical Expense Provisions, 26 Yale Law & Policy Review 309 (2007). Abstract

• Corporate Taxation
• Financial and Accounting Basics for Lawyers
• Income Taxation
• Property