Computers and Composition

Despite audience awareness being fundamental to writing pedagogy, student writers often struggle to move beyond considering their instructor as the primary reader, making connections between audience analysis and actual writing practice difficult. Generative artificial intelligence (GenAI) when viewed as models of culture that can be prompted to adopt various audience personas, may enable writers to interact with simulated audiences across multiple theories of audience consideration. We thus critically review a taxonomy of five audience theories: (1) writing without audiences, (2) specific people as audiences, (3) discourse communities, (4) publics, and (5) networks/algorithms. Rather than using GenAI as a text production tool, we argue for using these technologies as objects of critical inquiry that allow students to “craft and design” audiences through iterative interaction. This approach encourages writers to critique the biased design of the technology itself as well as their own biases embedded in audience descriptions.

This case study investigates how two ESL graduate students, Ian and Sam, use ChatGPT in their research writing after receiving a comprehensive tutorial based on Warschauer et al.’s (2023) AI literacy framework. We analyzed their engagement with ChatGPT across prompt categories including genre, content, language use, documentation, coherence, and clarity. Data were collected from research paper drafts, ChatGPT chat histories, and interviews. Data analyses included coding ChatGPT prompts, textual analysis of drafts, and thematic analysis of interview transcripts . Results show that while both participants utilized ChatGPT for understanding genre conventions and content development, they developed distinct approaches reflecting their individual backgrounds. Ian selectively used ChatGPT for specific assistance needs, while Sam engaged more systematically, particularly for APA style and coherence checks. Both approaches maintained academic integrity and scholarly voice, demonstrating that Generative AI tools can be effectively tailored to individual needs without compromising ethical standards. This study highlights how advanced ESL writers can adapt GenAI tools to their unique writing processes, offering insights into the diverse ways AI can enhance academic writing while preserving individual agency. The findings suggest that AI integration in academic writing can be customized to support diverse writing goals and backgrounds.