How Arbitrators Are Harnessing Artificial Intelligence

How Arbitrators Are Harnessing Artificial Intelligence
Posted on: Tue, 02/20/2024

Fast, fair, and full of potential, artificial intelligence (AI) tools can be used by arbitrators and mediators to boost efficiency, offer deeper insights, and provide an enhanced level of precision in their work. Read on to see how.

Streamlining Document Analysis

Generative AI tools, trained on vast data sets, have exceptional capacity for natural language processing, allowing them to rapidly search, compare, summarize, and extract insights from large volumes of text, images, and data, as follows:

  • Indexing and searching exhibits to instantly locate a fact cited in expert testimony 
    This eliminates manually combing records as well as delays requesting page numbers during hearings.
  • Identifying deposition discrepancies by analyzing multiple transcripts
    AI can automatically surface contradictory evidence across witnesses to spotlight potential issues.                       
  • Summarizing key details distilling critical information from lengthy briefs, such as pertinent procedural history from a preliminary-injunction hearing transcript  
  • Digesting document sets to generate visual timelines of relevant case dates and events, a task otherwise requiring manual compilation
  • Concept-based searching that goes further than simply matching keywords and can interpret the underlying concept of a search query

Streamlining Text Drafting

In addition to analytical power, platforms like ChatGPT and Claude can generate human-like text. While not legally foolproof, they allow professionals to craft customized drafts faster by:

  • Automating procedural and case-management templates based on specifics provided by the user, such as party numbers and dispute topics,
  • Producing initial draft settlements or uncontested arbitration awards based on prescribed terms, and
  • Standardizing repetitive components across documents, such as mediation brief boilerplate recitals.

As the technology progresses, the future could include

  • Development of predictive analytics tools forecasting how an arbitrator may rule on certain arguments based on historical patterns. Currently, scarcity of public arbitration data currently limits viability.
  • Speculative “soft” insights like gauging real-time emotions of parties through voice and facial analysis to help navigate discussions.

Three AI Platforms for Arbitrators and Mediators 

Here is a snapshot of three platforms utilized in the legal industry. Please note that these platforms are ever evolving with added capabilities.

CoCounsel (Casetext): specializes in legal research

  • Built-in database for legal research across case law, statutes, regulations
  • Shepardizes cases and links to source decisions 
  • Summarizes documents at adjustable detail levels
  • Compares similarities/differences in document contents

ChatGPT (OpenAI): has customizability and breadth

  • Answers questions by generating human-like text 
  • Analyzes submitted documents to provide summaries, comparisons
  • Researches legal issues using web data (risks outdated results)
  • Drafts customized legal documents based on user prompts
  • Iterates outputs based on user feedback

Claude (Anthropic): offers concise and practical applications 

  • Summarizes documents concisely, focusing on key details
  • Compares positions/arguments across document sets
  • Researches legal issues (notes limitations of training data)
  • Drafts legal text tailored to parameters provided by user
  • Refines content per user guidance

It must be emphasized that across current and future applications, AI is an enhancement rather than replacement for the specialized expertise of human arbitrators and mediators. When thoughtfully balanced, generative technologies can amplify their judgment, efficiency, and insights.


The above post is adapted from Dispute Resolution Enhanced: How Arbitrators and Mediators Can Harness Generative AI, an article by David L. Evans, Stacy Guillon, Ralph Losey, Valdemar Washingon, and Laurel G. Yancey. To read the article in its entirety, please click here. You will be directed to the AAAi Lab of the AAA®.