Technology-Assisted Literature Reviews with Technology of Artificial Intelligence : Ethical and Credibility Challenges

General discussion around artificial intelligence (AI) seems to be mostly concerned about potential benefits or hazards. While AI excels in automating basic and simple repetitive tasks, this is not a synonym for creativity, nor ethically sustainable unless the role is simple. Machine learning, deep learning and AI are trendy phenomena. However, the ethicality and credibility of the usage of AI in Technology-assisted literature reviews are raising questions. A systematic literature review was conducted to the SCOPUS database and Google Search. The PRISMA framework was used for the literature review. It was explored how AI is used to conduct -or assist in literature reviews. The results indicate that different scientific fields discuss issues relating to biases in AI technology and there is no consensus about what keywords and definitions should be to describe the same issue. AI has its place in the methodology of literature reviews, and this methodology continues to develop, but there are unsolved issues that how to create a tool with minimal bias and understand the current limitations of the technology of artificial intelligence. The empirical observation is that there are no specified keywords for the articles which relate to artificial intelligence-assisted literature reviews, which creates limitations for the generalisability of the results of this systematic literature review. Thus, this study has made several basic discoveries during this systematic literature review and their results are discussed. AI still poses bias risk and AI assisted literature reviews cannot be reliable without human supervision and there is no consensus on how AI assisted literature reviews should be classified in keywords and titles. It remains an open question whether there are any more hidden qualitative hypotheses in the development of reliability and validity of AI technology, which are waiting for discovery.