# OpenAI Cribbed Our Tax Example, But Can GPT-4 Really Do Tax?

arXiv: https://arxiv.org/abs/2309.09992 · HTML: https://mlsanity.com/paper/2309.09992/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Sep 2023.

## Built on by (most cited first, 4 of 4)

- [Evaluating AI for Law: Bridging the Gap with Open-Source Solutions](https://mlsanity.com/paper/2404.12349/) (2024)
- [BLT: Can Large Language Models Handle Basic Legal Text?](https://mlsanity.com/paper/2311.09693/) (2023)
- [DeonticBench: A Benchmark for Reasoning over Rules](https://mlsanity.com/paper/2604.04443/) (2026)
- [Language Models and Logic Programs for Trustworthy Tax Reasoning](https://mlsanity.com/paper/2508.21051/) (2025)

## Abstract (arXiv)

The authors explain where OpenAI got the tax law example in its livestream demonstration of GPT-4, why GPT-4 got the wrong answer, and how it fails to reliably calculate taxes.