Abstract
<title>Abstract</title> <p>The rapid proliferation of artificial intelligence tools in academic research has transformed how researchers search for literature and produce scientific writing. This article synthesizes current evidence on the use of AI tools, including Elicit, ChatGPT, ResearchRabbit and Consensus, across the scholarly workflow from literature discovery and systematic review screening to manuscript drafting and peer review assistance. Drawing on empirical studies, meta-analyses, comparative evaluations and bibliometric analyses published primarily from 2020 onward, this review examines the efficiency gains, quality improvements and ethical concerns associated with AI integration in scientific practice. Findings indicate that AI tools substantially accelerate literature screening and reduce manual workload, with some tools cutting screening time by up to 50%. AI-assisted writing tools improve textual cohesion and clarity, with documented benefits for non-native English speakers. The evidence also underscores persistent limitations, including algorithmic bias, reliability gaps, risks of overreliance and significant ethical challenges regarding transparency, authorship and academic integrity. The consensus in the literature favors hybrid human-AI approaches, in which automation complements rather than replaces expert judgment. Future research should prioritize standardized evaluation frameworks, domain-specific ethical guidelines and explainable AI models to ensure responsible integration of AI in scientific workflows.</p>