Generative AI: Research Use, Copyright and Ethics

Trustworthiness/Reliability: 

While many of these tools are drawing on vast amounts of reliable data, the black-box nature of these applications can at times make it difficult to verify the source of the the information they generate.  When researchers are unable to verfiy the source of the information and contents these tools produce, and the sources it was trained on can make it difficult to rely on the accuracy of the output.

• Consider consulting with a librarian who can work with you to help fact-check and verfiy information sources. 
• Consider consulting multiple sources of information in order to verify the outputs you are yeilding from generative ai tools.

Generative AI is NOT a research database:  

While these tools can be used for discovery, they are not a scholarly literature database nor are they a search engine. In some cases these products will have been trained using data from scholarly resources, but unlike a discovery platform which will point you directly to the source literature, generative AI tools are designed to "generate" their own content or output.

• Consider searching for literature in actual research databases or discovery platforms like HOLLIS, when looking for new resources
• Consider consulting with a librarian or expert in your field or discipline who can help you to both identify relevant titles and help verify their authenticity and provenance
• Features like reliability and trustworthiness are commonly highlighted in relation to scholarly research resources. Understanding and verifying the quality of the source material in generative AI tools can commonly be a challenge.

Privacy and Bias and Other limitations: 

While some products and tools are open source others are commercial and commonly capture information about users, biases in these products are derived from the data sets used to train them. Generative AI tools commonly reproduce biases inherently found in the data sets they are trained on which can perpetuate both harm and misinformation.

Generative AI tools continue to evolve and change, with new products continually being created and introduced.  We can anticipate that some existing tools and functionalities will be discontinued, while others will develop new enhancements and versions.

As different generative AI products are trained on different large language models, the data ranges and accuracy of the included information will vary from tool to tool.  When searching for timely or current information, an AI tool will only be as good as the dates of coverage included in the data set it was trained on.

AI tools for research can help you to discover new sources for your literature review or research assignment. These tools will synthesize information from large databases of scholarly output with the aim of finding the most relevant articles and saving researchers' time. As with our research databases or any other search tool, however, it's important not to rely on one tool for all of your research, as you will risk missing important information on your topic of interest.

NOTE: The resources described in the table represent an incomplete list of tools specifically geared towards exploring and synthesizing research. As generative AI becomes more integrated in online search tools, even the very early stages of research and topic development could incorporate AI. If you have any questions about using these tools for your research, please contact a librarian.

This content is evolving and subject to change. Last updated: December 20, 2024

Copyright considerations when using GenAI tools

There remains significant legal uncertainty with the use of Generative AI tools and copyright. This is an evolving area and our understanding will develop as new policies, regulations, and case law become settled.  

• Input: The lawfulness of the use copyright-protected content in the training of AI models is a question currently being contested in courts and by policy-makers in many jurisdictions. There are several lawsuits originating from the US that allege Generative AI tools infringe copyright, and it remains unclear if and to what extent the fair use doctrine can be applied in the context of AI model development. In Canada, there also remains uncertainty regarding the extent to which existing exceptions in the copyright framework, such as fair dealing, apply to this activity.¹ A whitepaper and the results of its latest round of consultations on copyright and Generative AI was published by the Government of Canada in June 2024 with a goal of continuing fact-finding work to inform copyright policy.²
• Output: Authorship and ownership of works created by AI is unclear. Traditionally, Canadian law has indicated that an author must be a natural person (human) who exercises skill and judgement in the creation of a work.³  As there are likely to be varying degrees of human input in content generated, it is unclear in Canada how it will be determined who the appropriate author and owner of works are. More recently, the US Copyright Office has published the following guide addressing these issues: Copyright Registration Guidance for Works Containing AI-Generated Materials 

If you would like to use GenAI tools for content generation, consider the following before doing so: 

• Have an understanding that while you can use these tools to create content, you may not own or hold copyright in the works generated 
• Be mindful of what you input into tools: never input confidential information or significant portions of intellectual property you do not have the rights or permissions to. All content entered may become part of the tool’s dataset and may inadvertently resurface in response to other prompts 
• Review the terms of service of each tool: These terms will dictate use and ownership of input/output and they are subject to change without notice 
• Be explicit in how you have used these tools in the creation of your work. Keep a record of prompts and any IP you have used in the creation of output. Review attribution guidelines according to the style guide you are using.. If you are publishing your work, review any requirements or policies that address the use of generative AI tools in your research. These policies will indicate whether AI can be used and how the use of these tools should be disclosed.

Generative AI tools have introduced new challenges in academic integrity, particularly related to plagiarism.

Plagiarism is typically defined as presenting someone else's work or ideas as one's own. While a generative AI tool might not qualify as a "someone," using text generated from an AI tool without citing is still considered plagiarism because the work is still not the researcher's own. Individual policies for using and crediting GAI tools might vary from class to class, so looking at the syllabus and having a clear understanding from the professor is important.

A note about plagiarism detection tools:

A number of AI detection tools are currently available to publishers and institutions, but there are concerns about low rates of accuracy and false accusations. Because generative AI tools do not generate large amounts of text word-for-word from existing works, it can be difficult for automated tools to detect plagiarism. Georgetown does not currently use the AI detection feature of its plagiarism detection tool, Turnitin.

False Citations

Another area of academic integrity affected by GAI tools is that of false citations.

Providing false citations in research, whether intentional or unintentional, violates Malone's Academic Integrity** policy.  GAI tools such as ChatGPT have been known to generate false citations, and even if the citations represent actual papers, the cited content in ChatGPT might still be inaccurate.

Recommendations

• If GAI tools are only permitted to be used for topic development, in the early stages of research, you might not need to cite them at all, but it's still important to check with your professor first.
• If you are providing commentary or analysis on the text generated by a chatbot and are either paraphrasing its results or quoting it directly, a citation is always required. You can find more information on citing GAI tools on this guide's Citing AI page.
• If you are a researcher planning to publish in a journal, it is best to review that journal's policies on the permitted use of Generative AI tools.
• It's important to always look up citations and check to make sure they are accurate, and if you're citing information from that source, to cite the original source rather than ChatGPT or whichever GAI tool you're using.

** Malone Catalog - Section on “Integrity - Academic”

1. Plagiarism¹
   1. submitting as one’s own work, part or all of an oral or written assignment which is copied, paraphrased, or purchased from another source, including world wide web, generative AI tools, and other online sources, without proper acknowledgement of that source. In written assignments, using three or more words in succession from a source without quotation marks and proper acknowledgement can be considered plagiarism.  
2. Accomplice in Plagiarism
   4. Academic or Scientific Misconduct

1. misrepresenting attendance or reason for absence
2. using information or accepting aid which is not approved by the instructor; this may include but is not limited to
   1. 1. 4. using a generative AI tool when completing an assignment

Breaches of Privacy & Danger of Re-Identification

There are currently also multiple privacy concerns associated with the use of generative AI tools. The most prominent issues revolve around the possibility of a breach of personal/sensitive data and re-identification. More specifically, most AI-powered language models, including ChatGPT, require for users to input large amounts of data to be trained and generate new information products effectively. This translates into personal or sensitive user-submitted data becoming an integral part of the collection of material used to further train the AI without the explicit consent of the user. Moreover, certain generative AI policies even permit AI developers to profit off of this personal/sensitive information by selling it to third parties. Even in cases when clear identifying personal information is not entered by AI user, the utilization of the system carries a risk of re-identification as the submitted dataset may contain patterns allowing for the generated information to be linked back to the individual or entity.  

Given these issues, extensive downloading of Library materials to build AI training corpora is prohibited. Additionally, some Library content providers prohibit any amount of their content being used with AI tools

Recommendations

• Avoid sharing any personal or sensitive information via the AI-powered tools. 
• Do not download Library materials (i.e., articles, ebooks, infographics, psychographics, or other datasets) into AI as it is prohibited.
• Always review the privacy policy of the generative AI tools before utilizing them. Be cautious about policies that permit for the inputted data to be freely distributed to third-party vendors and/or other users.