SciSpace Academy
TutorialLiterature review › Responsible and ethical AI literature reviews

A Librarian's Guide to AI in Literature Reviews with SciSpace

Dr. Olga KozEnglishLiterature reviewResponsible and ethical AI literature reviews

TL;DR

1-min read

AI tools can make literature reviews faster, but speed only helps if researchers still develop the judgment evidence synthesis requires. In this tutorial, Dr. Olga Koz, Collegiate Librarian at Kennesaw State University College of Education and an evidence synthesis methodologist with the Interactive Research Methods Lab, uses SciSpace as a case study for integrating AI responsibly. It is aimed at librarians, faculty, research supervisors and graduate students who teach or carry out literature reviews.

Key takeaways

  • There is a core tension between AI efficiency and building genuine evidence synthesis skills, and both need attention.
  • SciSpace can serve as a literature review assistant at specific points, while human judgment should lead key decisions.
  • Librarians, instructors and supervisors can teach reviews with AI by framing it as an assistant within a structured search process.
  • Students learn best when they practise defining questions, building search strategies and verifying AI summaries against the sources, rather than outsourcing these skills.
  • In systematic reviews, AI can support searching, screening and data extraction, provided inclusion and quality decisions stay with humans.
  • Documenting how AI was used and reporting the process transparently keeps systematic and traditional reviews rigorous.
Apply it

To apply this, design one review exercise in SciSpace in which students compare AI summaries with the original papers and discuss where the tool's limits appear.

How to do it

  1. Define where SciSpace fits

    Decide at which points of the review SciSpace acts as an assistant and which key decisions human judgment must lead. This balances efficiency with real evidence synthesis skills.

  2. Practise questions and search strategies first

    Have students define their own questions and build search strategies before using AI. Practising these skills prevents them from being outsourced.

  3. Compare AI summaries with originals

    Design a review exercise in SciSpace where students compare AI summaries with the original papers. Discussing the differences shows where the tool's limits appear.

  4. Keep inclusion decisions with people

    In systematic reviews, use AI in literature reviews to support searching, screening and data extraction, but keep inclusion and quality decisions with humans.

  5. Document and report AI use

    Record how AI was used and report the process transparently. Documentation keeps both systematic and traditional reviews rigorous.

Try this on SciSpace

Frequently asked questions

How can librarians teach students to use AI for literature reviews?

Librarians can frame AI as an assistant within a structured search process. That means teaching students to define questions, build search strategies, check results against databases, and verify every AI summary against the source. Pairing AI exercises with explicit discussion of its limits helps students build information literacy instead of relying on the tool to think for them.

Is it acceptable to use AI in a systematic review?

AI can support tasks such as searching, screening and data extraction, provided the review keeps its methodological rigor. Reviewers should document how AI was used, keep human decisions for inclusion and quality assessment, and report the process transparently. Checking journal and institutional guidance before submission is also important, since expectations about AI use differ across fields.

Related tutorials