Proceed with Caution: AI as a Search Tool

Are you using AI to conduct academic research?  Or maybe just to find out about a topic relevant to your practice?

You’re not the only one!  Coffey (2024) reports some kind of AI is being used by 75% of researchers….even though 92% of them don’t trust it.

This distrust is not without reason, and its good to see.  There are some serious pitfalls to using AI to find factual information or even to find research articles.  However, there are also some helpful AI based tools that may support your information seeking.

Traditionally, academic databases and more recently, institution-hosted discovery tools have been the way to find published, peer-reviewed information.  However, since the development of generative AI tools, such as ChatGPT and CoPilot, information seeking in all its forms has evolved (Ayre et al, 2025).

Some major issues with AI generated search results are bias and misinformation.  Bias can occur due to the fact that generative AI learns from previous data (Williamson & Fernandez, 2025) and anyone, anywhere can be inputting that data.  A major source of misinformation are hallucinations, where AI conjures up information that is inaccurate or false or a citation that does not exist (Williamson & Fernandez, 2025).  Here at District Library Services, we have been asked to retrieve articles generated by AI searches which we have found to be a hallucinations, despite how real they look!

Along with these issues, many AI search tools do not make it simple to retrieve the citations you need to then locate the articles (Williamson & Fernandez, 2025).

Its good to remember that those traditional old academic databases are more complex for a reason.  They have boundaries around what they search so that the results they retrieve are of higher quality and more easily analysed for relevance (Williamson & Fernandez, 2025) having been mapped to subject headings.

So why then do people use AI for these type of searches?

Part of the reason is NLP or Natural Language Processing.  Developing a search strategy for an academic database can be time consuming and if you are not skilled in this area, lead to less than desirable results.  But AI tools allow you to search in natural language which means you can type in a whole question or sentence and it will bring up relevant results (Gorton, 2025).

One of the positive aspects of AI’s ability to learn is that it can provide results more tailored to the individual user (Gorton, 2025).  Some tools will even provide suggestions of additional searches you could undertake!

So what next?  Do we throw the baby out with the bathwater? AI is here to stay and there are some tools that have a reduced risk of use.

Monash Health have helpfully analysed these tools for you and even provide a handy matrix for you to use to decide which to use (Gorton, 2025).

Each tool has its pluses and minuses and its individual characteristics may influence which you choose, but Monash recommend Consensus.app and Semantic Scholar and suggest you steer clear of Copilot and ChatGPT amongst others (Gorton, 2025).

If you choose to use AI for your searching, please keep this sage advice from Williamson & Fernandez (2025) in mind "It should be emphasized that while AI search may function as a valuable time saver, for high-stakes searchers, it remains unreliable enough that it is inherently problematic if it is to be used as the only method of research."

So while it can be a useful tool, it is no replacement for a thorough search strategy put through a variety of academic databases.  Reach out to the library for assistance with this.  We can undertake a literature search for you or provide training so that you are more equipped to undertake your own.  You can contact us here.

Monash Health’s fantastic Literature Searching Guide can be accessed here.  Look for the AI search tools section at the bottom.

Ayre, J., Cvejic, E. & McCaffery, K. J., (2025). Use of ChatGPT to obtain health information in Australia, 2024: insights from a nationally representative survey. Medical Journal of Australia, 222(4), 210-212. https://doi.org/10.5694/mja2.52598

Coffey, L. (2024). Most researchers use AI-Powered tools despite distrust. Inside Higher Ed. https://www.insidehighered.com/news/quick-takes/2024/05/24/report-most-researchers-use-ai-tools-despite-distrusting-it

Gorton, C. (2025). Tech Showdown – AI Search Tools Special Issue. Journal of Health Information and Libraries Australia, 5(1), 5-8.

Williamson, J. M. & Fernandez, P. (2025). “Through the looking glass: envisioning new library technologies” academic search using artificial intelligence tools. Library Hi Tech News, 42(2), 1-5. https://doi.org/10.1108/LHTN-01-2024-0014.