Research often begins before we know the right terms to search. You may have a question, an unfamiliar topic, or an early idea, but not yet know which authors, concepts, or phrases will lead you to useful scholarship.
This fall, JSTOR is introducing AI-assisted dynamic search, a new way to explore those early questions. It considers the meaning and context of a search to identify related ideas and potentially relevant sources across JSTOR. It also shows how it interpreted the query, giving researchers more context as they decide what to explore next.
Dynamic search is designed to support discovery while keeping the researcher in control. It can suggest connections and vocabulary, but it does not evaluate sources or draw conclusions for you. Researchers remain responsible for reading the original work, assessing its relevance, and deciding how it informs their research.
A different way to begin a search
Keyword search remains useful when you know the terms or names you want to find. Dynamic search is intended for moments when those terms are still taking shape. A researcher might begin with a broad question, review related concepts and sources, then refine the question or continue with keyword search.
The experience complements JSTOR’s AI research tool, which helps researchers engage with an article or book chapter they have already found. Dynamic search supports discovery across JSTOR; the research tool supports closer work with an individual source.
A fall release focused on learning
We are introducing dynamic search at participating institutions where JSTOR’s on-platform AI features are enabled. Library administrators retain control over whether those features are available at their institution.
This fall release is a chance to understand how researchers use dynamic search, which parts help them explore a topic, and where the experience needs improvement. We’ll learn from use and feedback to enhance how we support research. Dynamic search is available at no additional cost to JSTOR institutional participants.
Help us shape what comes next
If you try dynamic search, we’d like to hear what helped, what felt unclear, and what would make the experience more useful. Feedback from researchers, students, faculty, librarians, and publishers will help inform how we develop the experience.
To share feedback or ask a question, contact JSTOR Support. If you’re reporting a specific result, sharing the query and a brief description of what happened can help us investigate
