Ask a chatbot for letters written by women during the French Revolution and you will get an answer in seconds. Whether you get a source you can read, cite, and trust is a different question. Generative AI tools are increasingly the first stop for students beginning primary source research, and they fail in ways that can look like success: a plausible citation that leads nowhere, or a summary drawn from training data rather than the source itself.

This online workshop examines where these tools genuinely help with primary source discovery and where they break down. We will work through live examples of prompting, verification, and the distance between a description of a source and the source itself. 

Participants will leave with strategies for using AI tools as a starting point in primary source research, and for checking what those tools return before citing it.