A pilot study at Kennesaw State University is shifting the debate on student AI writing from outcomes after the fact to what students are actually doing during composition. Using think-aloud protocols, researchers observed how undergraduate writers interacted with generative AI tools while drafting, aiming to clarify whether students are outsourcing writing versus using AI in ways that still reflect active decision-making. The reporting comes amid broader classroom AI controversies, including questions about overreliance, cheating, and how instructional practices shape student critical thinking. The study’s methodology is notable because it tracks the writing process itself rather than only final papers or self-reported behavior. Separately, higher education educators and researchers are examining how to translate AI guidance into classroom instruction without flattening pedagogy into automated output—emphasizing that teachers still need to manage learning design, academic integrity, and discourse. For faculty governance and academic leadership, the near-term takeaway is that evidence about AI use is becoming more granular, which can influence institutional AI-writing policy, faculty training, and assessment redesign.