Can I Use Large Language Models and Other AI (such as ChatGPT, Google Gemini, etc.) with ICPSR Data?

General Statement: Large language models (LLMs) may only be used to manage, process, or analyze data distributed by ICPSR if the LLM meets specific criteria regarding retention of user-supplied data and/or placement within a secure network.

Under ICPSR’s Bylaws, researchers are forbidden to distribute data or other materials we supply (apart from study-level metadata and related publications described below) to other members, organizations, or individuals without ICPSR’s written permission. This includes self-published datasets brought in under the ICPSR Terms of Use (check the study’s home page for licensing information). This document explains how this policy affects the ability of researchers to use LLMs to analyze data distributed by ICPSR. This policy does not apply to ICPSR metadata.

Learn more about ICPSR’s Policy on the Use of Large Language Models.