College instructors and English departments are recalibrating course design as large-language models proliferate on campuses, shifting from outright bans toward mitigation, pedagogy redesign and testing for AI use. Some faculty are exploring new assessment formats, in-class handwritten exams, oral defenses, and viva voce presentations to ensure learning outcomes remain assessable. Meanwhile, a separate cohort of scholars argues that detection technology and agreed frameworks could preserve analytic writing assignments: reliable AI-detection tools would let educators enforce or permit LLM use with confidence. The debate centers on whether detection is now robust enough to distinguish human work from machine-generated text. Higher education officials are experimenting with both restrictive and integrative approaches; the near-term focus is on assessment integrity, instructional redesign, and investing in policy and tooling to audit or authorize AI use.