Generative AI-Assisted Writing in Indonesian Language Education: A Systematic Literature Review of Process Pedagogy, Critical Literacy, and Academic Integrity

Abstract

Generative artificial intelligence (GenAI) has rapidly entered writing classrooms, yet its educational value depends on whether it supports or bypasses the intellectual work of composing. This systematic literature review synthesizes 20 peer-reviewed international journal articles published between 2022 and 2025 to identify how GenAI, educational chatbots, digital multimodal composing, and critical digital literacy can inform Indonesian language education. A transparent search and screening protocol was applied to journal articles concerned with language learning, writing development, feedback, multimodal composition, authorship, equity, and academic integrity. The thematic synthesis produced five findings. First, GenAI can assist brainstorming, genre modelling, language reformulation, and formative feedback, particularly when students remain responsible for planning and revision. Second, process-based designs are more educationally defensible than product substitution because they make decision making and authorship visible. Third, hallucination, bias, privacy, overdependence, and unreliable AI-detection systems require explicit critical AI literacy. Fourth, multimodal and narrative research indicates that technology should strengthen meaning, cultural voice, and audience awareness rather than reward surface polish. Fifth, equitable implementation requires flexible access, teacher mediation, and assessment that values evidence, reflection, and revision history. Based on these findings, the article proposes a five-stage HUMAN-AI writing cycle for Indonesian language classrooms: human grounding, purposeful prompting, evidence checking, authorial reconstruction, and reflective disclosure. The framework positions GenAI as a supervised dialogic resource, not a substitute author.