Skip to Main Content

ChatGPT and Generative AI Legal Research Guide

Introduction to Claude 3

What is Claude 3?

Claude 3 Opus is the most advanced model in Anthropic's Claude 3 family, setting new benchmarks across a wide range of cognitive tasks.

Claude 3 Opus is currently the best AI system for legal research, according to Anthropic. It is available to Claude Pro subscribers for $20 per month. Anthropic also provides free models in the Claude 3 family for general use.

How Does It Work?

  • Claude 3 Opus is a large language model that generates human-like responses by predicting the next word in a sequence based on its training data.
  • When prompted with a question or task, it leverages its extensive knowledge to generate the most relevant and coherent response.


  • The training data for Claude 3 Opus only extends to August 2023, so it may not include the most recent information or materials.
  • As with all generative AI systems, Claude 3 Opus can produce inaccurate or hallucinated information, requiring independent verification and fact-checking.
  • While highly capable, it is an assistive tool and not a replacement for human expertise.

Claude 3 Legal Research Skills

Claude 3, the latest AI model from Anthropic, has demonstrated impressive capabilities in legal research and memo writing. In a recent test, a legal research instructor provided Claude 3 with a fact pattern and asked it to analyze the issue, list relevant Texas statutes and cases, and generate a 1,500-word legal memorandum. Claude 3 performed remarkably well, especially when compared to other AI systems like ChatGPT-3.5, Lexis+ AI, and Microsoft Copilot.

Claude 3's performance was remarkable, especially when compared to other AI systems like ChatGPT-3.5, Lexis+ AI, and Microsoft Copilot. It demonstrated the following:

  • Generated 7 real, relevant cases with correct citations in response to the legal research question
  • Cited 12 cases in the final version of the memo
  • Produced a high-quality 735-word draft memo on the first attempt, which included:
    • Correctly cited cases using Bluebook format
    • A well-organized Discussion section that effectively applied the law to the facts
  • After being asked to expand the memo twice, delivered a comprehensive 1,446-word document with a thorough Discussion section