The nature of generative AI (GAI) content, and the increasing popularity of GAI tools have led to important copyright questions including defining copyright ownership of generative AI works and the rights of creators whose works have been used to train AI models. The following provides an overview of copyright considerations to keep in mind when using GAI tools. Due to the rapidly changing GAI copyright landscape, the information presented below will change over time.
The two main areas of copyright concern revolve around generative AI input and output.
Input: Generative AI relies on large amounts of data to build its models and generate outputs. Some GAI tools may have trained their systems by copying content without permission from the creators or copyright holders. This has raised copyright infringement concerns and has been the focus of several lawsuits.
Output: In Canada, copyright ownership of generative AI content is still uncertain which means that a creator may not be able to claim copyright protection for works they create using GAI tools. At the same time, it’s challenging to provide attribution to the original authors whose works were used to train GAI models.
Learn more
Seneca Resources
Some content in this section was adapted from University of Toronto’s Generative AI Tools and Copyright Considerations, and University of Waterloo’s Generative Artificial Intelligence. Note: ChatGPT was used to generate topic ideas for this section. Perplexity AI was used to find sources for background information.
GAI tools have terms of use that outline ownership and appropriate use of the tool. It’s important to review and understand the terms so that you are aware of your responsibilities and rights as a user. For example, some tools may claim ownership of the output, or make users responsible for copyright infringement that may result from using the tool.
Always ensure that you have permission from the copyright holder or have verified that you can use a piece of text, image, or other content as input in GAI tools. It’s also helpful to learn more about the data used to train the GAI tool to identify potential copyright infringement issues.
Try to use prompts / inputs that are in the public domain or released under an open licence since these materials can be used more freely. For materials with an open licence, such as Creative Commons, make sure to review the licence terms to understand permitted use of the material. For more information on using Creative Commons content in GAI tools, see Understanding CC Licenses and Generative AI. Similarly, consider using GAI tools that only include training data that are from the public domain and/or the tool's developers have licensed.
Consider using original prompts and inputs instead of copying or adapting another creator’s work. Before using your original work, it’s important to review the tool’s terms of use to understand how inputs are handled within the tool.