Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
162 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Better Call GPT, Comparing Large Language Models Against Lawyers (2401.16212v1)

Published 24 Jan 2024 in cs.CY and cs.CL

Abstract: This paper presents a groundbreaking comparison between LLMs and traditional legal contract reviewers, Junior Lawyers and Legal Process Outsourcers. We dissect whether LLMs can outperform humans in accuracy, speed, and cost efficiency during contract review. Our empirical analysis benchmarks LLMs against a ground truth set by Senior Lawyers, uncovering that advanced models match or exceed human accuracy in determining legal issues. In speed, LLMs complete reviews in mere seconds, eclipsing the hours required by their human counterparts. Cost wise, LLMs operate at a fraction of the price, offering a staggering 99.97 percent reduction in cost over traditional methods. These results are not just statistics, they signal a seismic shift in legal practice. LLMs stand poised to disrupt the legal industry, enhancing accessibility and efficiency of legal services. Our research asserts that the era of LLM dominance in legal contract review is upon us, challenging the status quo and calling for a reimagined future of legal workflows.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (11)
  1. Amazon. Amazon titan: A large language model. https://aws.amazon.com/bedrock/titan, 2023. Accessed: 2023-12-19.
  2. Palm 2 technical report. arXiv preprint arXiv:2305.10403, 2023.
  3. P. D. Callister. Generative ai and finding the law. Available at SSRN 4608268, 2023.
  4. Chatgpt goes to law school. Available at SSRN, 2023.
  5. L. J. Cronbach. Coefficient alpha and the internal structure of tests. psychometrika, 16(3):297–334, 1951.
  6. Legalbench: A collaboratively built benchmark for measuring legal reasoning in large language models, 2023.
  7. A. of Corporate Counsel. 2023 law department compensation survey. https://www.acc.com/sites/default/files/2023-09/ACC_2023_Law_Dept_Compensation_Survey_Exec_Summary.pdf, 2023. Accessed: 2023-12-01.
  8. OpenAI. Gpt-4 technical report. arxiv 2303.08774. Open AI, 2:13, 2023.
  9. Chatgpt as an artificial lawyer? Artificial Intelligence for Access to Justice (AI4AJ 2023), 2023.
  10. A. Team. Claude: A large language model. https://www.anthropic.com/claude, 2023. Accessed: 2023-12-19.
  11. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288, 2023.
Citations (11)

Summary

  • The paper finds that LLMs match or exceed human accuracy while drastically reducing review time in legal contract analysis.
  • The study employs empirical analysis comparing precision, recall, and F-score metrics between LLMs and traditional legal professionals.
  • The paper highlights significant cost savings, with LLMs reducing expenses by up to 99.97%, transforming legal contract review workflows.

Introduction

Artificial Intelligence has been steadily making inroads into various professional fields, and the legal sector is no exception. The ability of LLMs to analyze, interpret, and generate human-like text has significant implications when applied to the intricate domain of legal contract review. This paper presents a comprehensive experimental analysis comparing the performance of LLMs against Junior Lawyers and Legal Process Outsourcers (LPOs) — the former generally being less experienced legal professionals and the latter frequently involved in routine legal tasks.

Benchmarking AI in Legal Contract Review

Central to this research is a trio of queries regarding the comparison of LLMs against human legal professionals in accuracy, speed, and cost-efficiency. Empirical results consistently showcase the prowess of LLMs, matching or even exceeding human accuracy in the identification of legal issues within contracts. In terms of speed, the reduction in time for LLMs to review contracts is staggering — the processing is accomplished in seconds versus the hours human counterparts require. On the financial spectrum, LLMs demonstrate a jaw-dropping cost reduction of 99.97 percent in comparison to traditional contract review methods.

Performance Metrics and Implications

Drilling down into the numbers, the LLMs exhibit compelling performance metrics. Models such as GPT4-1106 displayed precision, recall, and F-score values closely rivaling those of experienced legal practitioners, with the bonus of a dramatic reduction in time taken for contract review. Moreover, when the economic aspect was evaluated, costs saving was evident, with figures like $0.02 to$1.24 per document for LLMs compared to $74.26 for Junior Lawyers and$36.85 for LPOs. These numerical results are a strong indicator of LLMs' potential to revolutionize how contract review is traditionally undertaken.

Future Outlook and Considerations

The implications for the legal profession are profound. This research doesn't only provide a static comparison but also hints at a dynamic shift in the roles and tasks of legal practitioners as LLMs continue to advance. The insights gathered suggest that LLMs can take on the burden of high-volume, routine legal tasks, freeing up human lawyers to tackle more complex issues thus enhancing the profession's overall value proposition.

Moving forward, it's clear that LLMs will significantly change the landscape of legal contract review. As they become more integrated, they stand to not only augment the capabilities of legal professionals but also democratize access to legal services due to their cost-effectiveness. However, this remarkable potential comes with the imperative to continue refining the technology, ensuring that it aligns with ethical standards and maintains the delicate balance between automation and human oversight in the legal field.

Youtube Logo Streamline Icon: https://streamlinehq.com