Scientific Research and Communication — AI Tools for Research

New tools are constantly emerging from LLMs (Large Language Models) that use generative AI to produce new analysis, text, or to answer questions.  Below I have listed some AI tools that are especially developed to aid with scholarly research.  Please note that listing these tools does not imply that I encourage or endorse them. I have listed the UConn Compliance rules with regards to AI in the next section of this page.  Please be aware of the risks and consequences of uploading information to AI tools with regards to either:

• sensitive or confidential information (DO NOT under any circumstance upload sensitive, confidential, or proprietary information to these tools)
• copyright infringement (many publishers/journals/authors prohibit the uploading of articles or their writing to LLMs)
• data ownership policies (read the fine print)
• AI hallucinations. An AI hallucination is incorrect or misleading information generated by an AI model. AI tools can hallucinate references, resources, and links. They can make up references that look legitimate, but are false or do not exist.  Do not blindly trust references given by chat prompt LLM AI tools like Chat GPT. You must verify the references given by a LLM against a legitimate library academic database citation.  

That being said, we now live in the era of ubiquitous AI. Many of these tools developed specifically for researchers can be very useful.  Best practices with regards to using AI include acknowledgement of the use of an AI tool in your work, citation, and checking for errors,

Make sure you A.C.E. your use of these tools by:

• Acknowledging that you used a LLM AI tool in your research or assignment
• Citing the tool in your bibliography or list of Works Cited
• Evaluating for errors: check for errors in references; is the information given by the AI tool true? Does it add value?

This list was created by and is maintained by STEM librarian Renée Walsh. Renée created the acronym A.C.E. as a pedagogical mnemonic device. It can be reused and remixed by attribution under a CC-BY Open Commons license.

Remember to Acknowledge, Cite, and Evaluate! A.C.E.

Are you interested in professional development about generative AI from UConn?  Please see the link below to UConn CETL workshops as well as a link to the UConn BEACON AI club.

According to the UConn Office of Compliance:

UConn has two AI services available to employees – Microsoft 365 Copilot and Microsoft 365 Copilot Chat. These two services are employees' safest data privacy options when using an online AI tool.  Both are fully secure under UConn's agreement with Microsoft, which means they are safe to use with UConn's data. As long as you are signed in with your UConn credentials, your prompts, the information retrieved, and the responses generated are not used to train the underlying AI models.

Alternative AI tools like ChatGPT should not be used in your work at UConn. Any protected or confidential information shared with these AI tools would be used to train the underlying AI models and publicly available. The public disclosure of UConn-protected or confidential information would be in violation of UConn policy. 

The university library research guides, webpages, and Open Access books listed below may be useful to you: