Guidelines for Using GenAI in Systematic Literature Reviews
A new preprint on arXiv (2607.24991) presents preliminary guidelines for using and evaluating generative AI (GenAI) and large language models (LLMs) in systematic literature reviews (SLRs). The authors note that while GenAI can summarize text, it may not meet the rigor, reliability, and transparency required for SLRs. Their methodology included a rapid review of existing guidelines, thought experiments, and their own experience. The paper identifies problems researchers face when evaluating GenAI for SLRs and offers recommendations for use and assessment. The work aims to support researchers conducting SLRs with GenAI or studying how well GenAI supports SLR tasks.
Key facts
- Preprint on arXiv with ID 2607.24991
- Focuses on GenAI and LLMs for systematic literature reviews
- Highlights potential lack of rigor, reliability, and transparency
- Methodology includes rapid review, thought experiments, and experience
- Identifies problems in evaluating GenAI for SLRs
- Provides recommendations for using and assessing GenAI
- Targets researchers conducting SLRs with GenAI
- Also supports empirical studies on GenAI's effectiveness for SLRs
Entities
Institutions
- arXiv