This information was last updated December 2024 and is subject to change in the future so please consult frequently and contact the library if you have questions [copyright@durhamcollege.ca].
What do I need to know about copyright and generative AI tools?
As a faculty member at Durham College, it is your responsibility to determine and communicate the authorized or prohibited use of GenAI in the completion of any academic work or assessments
As a student at Durham College, it is your responsibility to familiarize yourself and comply with the guidelines for using GenAI in each of your classes, because each instructor may have different permissions and procedures.
For all users, it is important to understand that both the input and output of GenAI is still being explored in copyright law.
- Input: Use of content to train AI models is currently being explored in lawsuits originating from the US. These cases allege that genAI tools infringe copyright and in Canada it is similarly unclear thus far the extent to which existing copyright exceptions, such as fair dealing, might apply.
- Output: Authorship and ownership of works created by AI is also unclear. Traditionally in Canada copyright can only be granted to a natural person or human author, who can exercise skill and judgement in the creation of a work. With varying degrees of human input involved in generating content with AI, it is not yet clear who the legal author and owner of works will be in Canadian law.
Can I input the full-text of library licensed resources, like journals or eBooks, in generative AI tools?
In most cases no. The library signs contracts with vendors and publishers and agrees to terms for all users, including students and faculty, which you are responsible for complying with when you access content. Most agreements do not allow for uploading licensed content to third-party platforms, including GenAI tools.
As AI develops rapidly best practices do as well. Before using library licensed content in an AI tool, consult the library for support to find out if the license for a given resource may permit the use or training of licensed content in public versions of AI tools.
Durham College's protected enterprise edition of Microsoft Copilot is an exception to this as long as you have used your DC credentials for access, as content is not saved, used to train models, or made available to Microsoft.
What other considerations are there for content generation with GenAI tools?
- Understand ownership of outputs: while you might use these tools to create content, you may not own or hold copyright to the works generated
- Be aware of inputs: never input confidential information or intellectual property you do not have rights or permissions for in Generative AI. Do not input library-licensed resources or student work. All content entered can become part of the tool's dataset and invertedly resurface in response to other prompts.
- Check the terms of use: Review the terms of service for each tool, which dictate use and ownership of inputs and outputs and are subject to change.
- Acknowledge use: Keep a record of prompts and any materials you input, and be explicit about how you used these tools in the creation of your work. Review attribution guidelines for the citation style you are using. If you are publishing your work review any requirements or policies of the publisher for addressing use of GenAI. These will indicate whether and how GenAI can be used and the use disclosed in your research. If you are a student, review GenAI allowance and attribution policies for course work in each class.
