Skip to Main Content
Eastern Michigan University Halle Library

Generative AI

This guide walks you through the considerations of using generative artificial intelligence (gen AI): selecting gen AI tools, best practices for using them, ethical considerations, and more.

How do I know gen AI results are accurate? 

You don't! Generative AI results need to be evaluated for accuracy and bias as much, if not more, than other sources. False or inaccurate gen AI results are known as hallucinations. Here are some tips for determining if the results are accurate: 
  1. Verify the source of the information: Find the source of the information and verify that it says what AI says it says. Some tools will give you a source for the information stated in the output, or you may have to ask the tool to give you a source. Locate that source of information and read it to see if it actually has the information the gen AI tool output said that it did. 
  2. Evaluate the original source: Is the information in the original source reliable? Are the authors listed and do they have authority on the topic? What is the bias of the source? Is the source peer-reviewed? 
  3. Cross check: Find other reliable sources to see if they agree with the information being provided in the AI output. 

More information about evaluating sources can be found at: https://guides.emich.edu/help/evaluating .

Read more about hallucinated citations on the Nature news article: "Can researchers stop AI making up citations?" by Elizabeth Gibney, Sept. 8, 2025. 

Recommendations from Cochrane for Using AI for Evidence Synthesis

If using AI for evidence synthesis, some considerations are necessary to ensure that the information created can be reproduced and evaluated. These recommendations are provided by Cochrane, from "Cochrane's approach to responsible AI use in evidence synthesis," by Ella Flemyng, Head of Editorial Policy & Research Integrity, Cochrane and Fabio Di Bello, Senior Customer Education Manager, Wiley. 

1. Remain ultimately responsible for the evidence synthesis
2. Report AI use in your evidence synthesis manuscript transparently
3. Ensure ethical, legal and regulatory standards are adhered to when using AI
4. Contribute to the ecosystem to help all roles continue to develop and grow