This paper outlines a teaching intervention for tertiary-level engineering students that incorporates Generative Artificial Intelligence (GAI) tools into an existing five-step academic writing framework. In alignment with newly developed institutional guidelines on AI use—similar to policies at many universities—the approach adapts Ethan Mollick’s recommendations on when (and when not) to employ AI. By embedding AI support into each stage of the writing process (pre-drafting, drafting, revising, editing, proofreading), the intervention aimed to preserve academic integrity while fostering student agency.

While the framework shares conceptual overlap with the SPACE framework (Set directions, Prompt, Assess, Curate, Edit), the primary emphasis remained on systematically integrating AI into this established academic writing pedagogy and augmenting an already robust framework for students to compose texts according to. Students were introduced AI tools such as Consensus, Scite_, Humata, Notebook LM and others as a semi-structured qualitative small scale investigation of student use of AI tools in the early part of the 2024/25 academic year had revealed that the surveyed students (N = 9) were using only Chat GPT. Furthermore, they were using it for purposes perhaps contrary to its intended use as an LLM, such as solving advanced mechanics and strength of materials problems. This suggested a low degree of GAI literacy on the part of the students, which had not been presupposed a priori.

The idea was to encourage students to use the GAI tools responsibly, with explicit attention to university guidelines, in particular ethical considerations, data privacy and the risk of deskilling, while at the same time increasing their writing efficiency. Indeed, it transpired that prompt-engineering needed to be explicitly addressed as an auxiliary technique, although this was secondary to guiding students in applying AI judiciously within the five-step model.

By documenting how Mollick’s “when-to-use” criteria enhance this academic writing framework—and by introducing AI tools beyond students’ initial scope of knowledge—this paper hopes to promote a replicable, student-centred framework for integrating GAI into the process of creating academic compositions and thereby improve, and not undermine, essential writing skills.

Keywords: Academic Writing, Generative AI, Engineering Education, AI Literacy, Deskilling, Academic Integrity.