Artificial Intelligence (AI) software can generate novel solutions to problems, including generating responses to user queries using a variety of methods, such as using keywords and phrases that were input by you into the search prompt. Popular generative artificial intelligence (GenAI) and chatbot models are built on large language models (LLMs) that are a type of artificial intelligence trained to look for patterns in large data sets.  However, this doesn’t necessarily mean that the information the AI tool provides is correct or from a reliable source. If the resources that the AI software provided cannot be found or accounted for, it can be difficult to rely upon the accuracy of the information. Therefore, further steps in fact-checking the information need to be conducted, such as consulting multiple different sources about the topic to determine whether the outputs generated are accurate.   Remember: AI is NOT a research database. While some of the information it presents may be pulled from scholarly resources, it cannot generate the actual source of that information.

When evaluating GenAI as research tools ask questions like: What is my method for checking the results? How does this AI tool use or access scholarly data? If you review the scholarly data directly, is what the AI tool states accurate?  Does it reliably point to real sources?  The Generative AI for Research guide expands on these questions with tips for evaluating GenAI tools. 

It’s best to use Yale Library's Database to search for scholarly journals and articles directly from our licensed resources. The library provides guidelines about research databases and AI usage with licensed materials. Please review the library's Using AI in Research page for specific project assistance.

If you decide to use AI to create content, you may not actually hold any copyright on the work that the resource generates for you.  When using AI tools, consider keeping a record of any prompts and IP that were put together that resulted in the final product. If this design is intended to be published, review all the publisher’s policies in regard to AI usage beforehand. Using works that you hold the copyright or works in the Public Domain with AI tools may be fine. Using other other people’s work (regardless of whether it’s full or a partial section) of intellectual property (IP) that you don’t own or have permission to use requires careful consideration. In using any third-party copyrighted content, always consider whether you can rely on Fair Use.

Please refer to Yale’s Rights Clearance Guide for Digital Projects  from the Office of General Counsel for more information about obtaining rights from third parties for content use and review Fair Use Tools to determine if you can make a case for fair use.

Additional information about copyright and AI in general can be found at the U.S. Copyright Office: Copyright and Artificial Intelligence, Part 1 Digital Replicas Report  followed by Part 2: Copyright and Artificial Intelligence, Part 2 Copyrightability Report and Part 3: Generative AI Training.

Copyright and citation are often conflated, but they are not the same and do not have the same requirements. In academia, citation is a critical part of scholarly practice. With increasing use of  AI in research, institutions, funders, and publishers are starting to require citation and documentation of its use. Documentation of scholarly AI use should be part of normal citation practices. Scholars should think about how they track using AI in research and publishing. Please review Reporting and Citation box on the AI for Research Guide for guidance.

Deciding to use Artificial Intelligence (AI) for your research can enhance your work in several different ways. It can add efficiency to the research process and workflow, it can enhance productivity, and it can help you brainstorm and create ideas for numerous different projects. However, there are several considerations to keep in mind when incorporating the use of these tools into your research process. 

Which AI tool is best to use? Which ones are safer than others? Which ones are free, and which ones require a subscription? You should also consider the content you want to use with an AI tool. Is there anything related to the following you should think about before proceeding?

• intellectual property

• cultural heritage

• privacy

• ethical considerations  

Yale has compiled a list with descriptions for each AI resource and how much risk each one imposes.  Yale’s AI Tools can be found here:  Yale’s AI Tools and Resources | AI at Yale while the data risk classification system Yale uses can be defined here: Data Classification Guideline (1604 GD.01)  

Additionally, the Scholarly Communication and Information Policy page provides excellent information on open access, author’s rights and publishing resources.

Generally, the answer is no. This is largely due to copyright and licensing agreements between the library and vendors that govern these resources.  

Yale has a License Review Steering Committee  that is dedicated to negotiating and reviewing electronic resources for the Yale Community. This pertains to all of the databases, music, and art that Yale offers its faculty, students, and staff.

The Subject Specialist in your area may be able to provide information on resources for your scholarly research. 

Please review our general guidelines which discusses whether a particular licensed e-resource can be used or not in association with AI.  For questions on existing licenses or about appropriate usage of licensed materials, please email LRSC.

If you are unsure on how to move forward with potentially using AI and licensed resources, or need copyright guidance in general, please contact Yale Copyright Librarian.

For legal advice in regard to using AI and publishing and other copyright infringement concerns, please consult with your own attorney or Yale's Office of General Counsel.