Generative artificial intelligence and critical digital literacy in EFL academic writing: An integrative review and pedagogical framework

Generative artificial intelligence (GenAI) has rapidly entered English as a foreign language (EFL) academic writing through tools that draft, paraphrase, translate, summarize, evaluate, and imitate disciplinary texts. The central question for language education is no longer whether learners will encounter these tools but how pedagogies can help them use AI critically, ethically, and rhetorically. This article offers an integrative conceptual review of peer-reviewed scholarship on AI-assisted writing, automated writing evaluation, digital literacies, identity, academic integrity, and EFL/ESL pedagogy. Drawing on studies from applied linguistics, language education, educational technology, and discourse studies, this paper synthesizes four recurring issues: AI writing systems as feedback infrastructures, learner agency and identity in human-AI composing, the risks of dependency and homogenized discourse, and the need for assessment practices that value process evidence rather than detection alone. The review argues that GenAI should be conceptualized as a literacy environment that mediates language, power, authorship and intercultural communication. It proposes a Critical GenAI Writing Literacy Cycle comprising six stages: orienting to task and genre, prompting strategically, comparing outputs, verifying evidence, transforming texts through human revision, and disclosing AI use through reflective accountability. The framework contributes to JLLI's scope of JLLI by connecting applied linguistics, technology-enhanced language learning, digital discourse, cultural studies, and language education. It concludes that responsible GenAI integration requires pedagogical designs that protect linguistic diversity, strengthen critical reading, and position EFL writers as accountable authors, rather than passive consumers of machine-generated prose.