Jump to content

Neel Guha

From AI Law Wiki

AI-generated text. This page was generated using artificial intelligence.

Scholarship · Legal AI evaluation and regulation

Neel Guha is a computer scientist and legal scholar at Columbia Law School. His work concerns AI's effects on legal practice, challenges for courts and regulators, and computational analysis of law. He joined Columbia as an associate professor of law on July 1, 2026.[1]

Roles and affiliations

Columbia's October 2026 account of its AI initiative describes Guha's participation in the incoming students' August orientation. His presentation addressed different types of AI tools and how to judge their suitability for particular legal tasks. The initiative combines teaching, research and discussion of AI's wider institutional consequences.[2]

Contributions and positions

Guha is a coauthor of LegalBench: A Collaboratively Built Benchmark for Measuring Legal Reasoning in Large Language Models. The project's original 2023 abstract describes 162 tasks covering six types of legal reasoning, developed with legal professionals. It presents the benchmark as a way to study which legal-reasoning tasks language models can perform.[3]

The authors describe aligning the tasks with legal frameworks so that lawyers and model developers can discuss evaluation using a shared vocabulary. The abstract reports an evaluation of 20 commercial and open-source models.[3]

Regulation and liability research

His faculty bibliography includes coauthored research on the feasibility of disclosure, registration, licensing and auditing in AI regulation, and on liability risks associated with health-care AI. Mariano-Florentino Cuéllar is among the coauthors of the regulatory-alignment article. These publications concern distinct legal questions alongside Guha's work on evaluation.[1]

Selected works

  • LegalBench (2023) — collaborative benchmark paper; original abstract and manuscript record.[3]
  • AI Regulation Has Its Own Alignment Problem: The Technical and Institutional Feasibility of Disclosure, Registration, Licensing, and Auditing (2024) — coauthored article listed in Columbia's bibliography.[1]

Coverage

References