Human–AI interaction in academic knowledge co-creation: a typology based on a systematic literature review
Przemyslaw Tomczyk, Demetris Vrontis, Moreno FrauPurpose
This study aims to examine how artificial intelligence (AI) contributes to knowledge co-creation in academic writing, focusing on systematic literature reviews (SLRs) as a representative and consequential form of structured knowledge production. It develops a typology of human–AI interaction that reveals how intellectual responsibility, decision-making and dialogue are distributed between human researchers and AI systems.
Design/methodology/approach
A systematic review of 120 publications from Scopus and Web of Science was conducted to analyse applications of AI in SLRs. Studies were coded along two dimensions derived from human–AI interaction theory: (1) decision-making delegation and (2) iterative dialogue between humans and AI. Using deductive thematic analysis, the research identified distinct interaction types and synthesised empirical patterns into a transparent framework.
Findings
The analysis reveals three types of human–AI interaction: AI-Supported, Collaborative and Partner-like. The AI-supported type treats AI as an assistive tool for automating repetitive tasks while preserving full human control. The collaborative type represents dialogical co-creation, where humans and AI engage in iterative prompting and feedback cycles. The partner-like type reflects hybrid partnership and partial delegation, where AI contributes analytical reasoning and interpretation under human oversight. Together, these types form a continuum of knowledge co-creation processes in which AI evolves from an efficiency tool to an epistemic partner.
Research limitations/implications
The dataset was limited to English-language sources indexed in Scopus and WoS, and coding relied on reported practices that varied in transparency. Future research should include experimental or longitudinal studies to assess how different interaction types affect review quality, transparency and ethical outcomes. The typology extends knowledge co-creation theory by including non-human (AI) agents as active participants in the SECI process and the dialogical model of co-creation.
Practical implications
The typology provides a framework for researchers, institutions and editors to assess how AI contributes to academic knowledge production. It supports responsible integration of AI in SLRs by clarifying levels of delegation and dialogue. The model encourages teams to document prompting cycles, maintain ethical oversight, and develop competencies for reflexive, transparent collaboration with AI systems.
Originality/value
To the best of the authors’ knowledge, this study is the first to conceptualise SLRs as sites of human–AI knowledge co-creation, linking knowledge management theory with evidence on AI-assisted academic writing. It proposes a theoretically grounded typology that clarifies AI’s role in distributed knowledge creation and offers practical guidance for transparent and ethical research practice.