Skip to sidebar after main content

Leadership: Using AI in Research

This page provides information, links, and tips to help guide you if you choose to use Artificial Intelligence tools as part of your research.

alt=""

AI is a great tool for some research tasks. If you are considering using AI as part of your research strategy, please explore this section carefully before you start.

Slide Deck: Using AI Tools in Your Research - a short presentation intended to help you think about AI and when it should or should not be used as part of your work.

Generative AI Guidelines - a general framework on using AI and its impacts on academic integrity from Wilfrid Laurier University.

Citing Use of AI - this Laurier Library webpage gives you a quick breakdown of how and when to cite AI when researching.

What kind of tasks is generative AI good for in academic research?

  • brainstorming
  • creating or refining a research question
  • generating keywords and subject terms for searches
  • finding webpages related to particular topics
  • spotting patterns or trends online
  • summarizing text found on the open internet
  • reviewing and editing your work (your own essay or project) for content and readability

alt=""   Never copy and paste an article, or part of an article or eBook, from one of the library's paid databases into a Large Language Model (LLM) AI. This is a copyright violation which could get you in a lot of trouble and could jeopardize the Library's license agreements with our database providers. The good news is, most databases are working on their own built-in AIs to help you with your research. Check the "Help" button on any database to see if AI tools are available. 

In addition to database-provided AI's, "academic focused" generative AI tools are being developed, which usually use open access materials. Click here to see a list of these types of AI. Laurier Library does not endorse use of any particular AI tool.

alt=""  Always check with your instructor to ensure use of AI is allowed in any particular assignment. Most assignments are created to help you learn how to do specific tasks on your own, and think critically about what you read, meaning AI won't be right for your work.

alt=""  Always check anything generated by AI for errors. AI can create false information and confidently state it as true. Ask the AI to provide citations and check them yourself to ensure they exist.

alt=""

TIP for open searches of the internet: be as specific as possible when asking a LLM AI to perform a task

  • give the AI a designation: e.g. "You are a professor of ___________ looking for academic resources about _______________"
  • provide context to the AI: e.g. what you want to prove or disprove, the purpose of the research, who will be the audience for the completed work
  • assign the AI the question or task: e.g. find journal articles or web pages about _____________ and provide a 3 sentence summary of each, and a citation for each in APA style
  • give the AI additional instructions or limitations: e.g. nothing published online before 2017, no newspaper articles, etc.

* based on a presentation at the OLA Super Conference 2024 by Avery Swartz, CampTech.ca

Prompts for analyzing documents without AI hallucinations

After you have located specific documents (not those found in the Library's databases or catalogue), help the AI to avoid hallucinations by giving it very specific guidelines when analyzing them. This only works with "high reasoning" AIs. Upload the documents into the AI and use the following prompts:

Base your answer only on the uploaded documents, nothing else.

If information isn't found, say "Not found in documents." Don't guess.

For each claim, cite the specific location - document name, page/section, and a relevant quote.

If you find something related but aren't fully confident it answers the question, mark it as [Unverified].

Only respond with information you are 100% confident is from the document.

[ Add your specific question here ]


Once you have an answer from the AI, ask it in the same thread to check its own work:

Re-scan the documents. For each claim, give me the exact quote that supports it. If you can't find a quote, take the claim back.
 

Upload the documents and the analysis provided by the first AI to another "high reasoning" AI, and ask it to check the first AI's work:

Review this analysis against the uploaded documents. Flag any claims that aren't directly supported.

* based on a video from Dylan Davis, YouTube January 31, 2026

Page Owner: CJ Nyssen

Page Feedback

Last Updated: