Using an Artificial intelligence chatbot to critically review the scientific literature on the use of Artificial intelligence in Environmental Impact Assessment

Abstract

There is considerable uncertainty about the role that Artificial Intelligence (AI) might play in Environmental Impact Assessment (EIA), including into research. AI large language model (LLM) chatbots have the potential to increase the efficiency of EIA research, but their outputs can create concerns. This paper investigates the potential time savings achievable using LLM chatbots to undertake a critical review of literature focussing on the use of AI in EIA. Using a combination of ChatGPT and Elicit, literature was reviewed to identify 12 key issues associated with the use of AI in EIA and this paper was prepared in three and a half days from initial conception. A protocol is developed to assist researchers in fact checking evidence delivered through Elicit (or other machine learning tools) which serves as a novel outcome of this research. Using comments from three peer reviewers allowed some more objective reflection on the credibility of the LLM chatbot-derived output, on the appropriateness of the time savings, and on the future research needed on the application of LLM chatbots in this context.