Are you wondering whether you can upload library books or articles, or otherwise mine or work with library content, in an AI system like Claude or ChatGPT? The answer is, it depends.
The library's licenses with publishers and vendors describe what can and cannot be done with library content, including AI use. What you can do differs depending on who owns and licenses the content rights. Your particular use case will make a difference as well. Fair use governs the use of all copyrighted materials, library or otherwise.
To learn how you may or may not use something (like a book, journal, video, or dataset) find the item or resource in Scholar OneSearch.
Then, click on "Show License".
In the resulting view, the license is summarized.
Look for sections labelled AI Use Allowance and/or AI Use Allowance Note. This will tell you what AI uses are permitted and prohibited under Northeastern's license.
As in the example below, the library's license with a publisher may specify that the content cannot be used to train LLMs, or the AI tool where you upload it must be a closed environment limited to authorized users. If you are logged into Claude.ai with your Northeastern account, you are in a closed environment, and your inputs are not used to train the model (see more about Claude).
For help interpreting the license, please email the library. For questions about responsible use of AI consistent with university policies, contact Northeastern's AI Review Committee.
- Note: If you're seeking an AI system for academic researching and summaries, the library recommends scite.ai or Scopus AI Assistant.
These two generative AI systems attempt to answer academic questions and summarize research in a user-friendly, transparent way, with real scholarly sources you can access and read. Many other library databases and search services also provide AI support. See this FAQ for more information.