Can LLMs Identify Tax Abuse?
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908780288016384 |
|---|---|
| author | Blair-Stanek, Andrew Holzenberger, Nils Van Durme, Benjamin |
| author_facet | Blair-Stanek, Andrew Holzenberger, Nils Van Durme, Benjamin |
| contents | We investigate whether large language models can discover and analyze U.S. tax-minimization strategies. This real-world domain challenges even seasoned human experts, and progress can reduce tax revenue lost from well-advised, wealthy taxpayers. We evaluate the most advanced LLMs on their ability to (1) interpret and verify tax strategies, (2) fill in gaps in partially specified strategies, and (3) generate complete, end-to-end strategies from scratch. This domain should be of particular interest to the LLM reasoning community: unlike synthetic challenge problems or scientific reasoning tasks, U.S. tax law involves navigating hundreds of thousands of pages of statutes, case law, and administrative guidance, all updated regularly. Notably, LLM-based reasoning identified an entirely novel tax strategy, highlighting these models' potential to revolutionize tax agencies' fight against tax abuse. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_20097 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Can LLMs Identify Tax Abuse? Blair-Stanek, Andrew Holzenberger, Nils Van Durme, Benjamin Computational Finance Artificial Intelligence I.2.1; I.2.7; J.4 We investigate whether large language models can discover and analyze U.S. tax-minimization strategies. This real-world domain challenges even seasoned human experts, and progress can reduce tax revenue lost from well-advised, wealthy taxpayers. We evaluate the most advanced LLMs on their ability to (1) interpret and verify tax strategies, (2) fill in gaps in partially specified strategies, and (3) generate complete, end-to-end strategies from scratch. This domain should be of particular interest to the LLM reasoning community: unlike synthetic challenge problems or scientific reasoning tasks, U.S. tax law involves navigating hundreds of thousands of pages of statutes, case law, and administrative guidance, all updated regularly. Notably, LLM-based reasoning identified an entirely novel tax strategy, highlighting these models' potential to revolutionize tax agencies' fight against tax abuse. |
| title | Can LLMs Identify Tax Abuse? |
| topic | Computational Finance Artificial Intelligence I.2.1; I.2.7; J.4 |
| url | https://arxiv.org/abs/2508.20097 |