A new study in *Computers & Composition* suggests that using generative AI for writing can shift how students think—pushing more attention to higher-order decisions rather than sentence-level drafting. Researchers led by Iowa State’s Abram Anders and Embry-Riddle’s Emily Dux Speltz studied students in a semester-long “AI and Writing” course. Across the term, students collaborated with AI chatbots to draft and then evaluate outputs, including identifying errors. The researchers report that while students may spend less time turning ideas into paragraphs, they invest more effort in planning, judgment, evaluation, and quality control—work aligned with critical writing practices. The study also connects to adoption signals: the article notes Gallup/Lumina polling that shows more than half of college students use AI weekly, while a large majority of professors report attempting to “AI-proof” assessments. The classroom findings feed into what AI-reliant instruction may look like if educators redesign tasks around verification and revision rather than banning the tool. For faculty and program leaders, the immediate operational takeaway is assignment design: if courses require prompt refinement, source- or output-checking, and iterative resubmission, AI can become a structured component of learning rather than a cheating vector.