Business schools are drawing sharper lines around how AI changes learning assessments and what counts as credible evidence of student work. One report describes an Ivy League professor using ChatGPT to test a take-home exam after suspiciously high scores in a Brown University math class, concluding that take-home exams may no longer reliably measure learning. Separately, the AI-era classroom is also pushing schools to reconsider what “teaching” looks like when tools can generate plausible outputs. The underlying governance question is assessment integrity: institutions may need new proctoring approaches, new assignment designs, and clearer policies for student AI use. For higher education leaders, the near-term impact is immediate policy work across faculty governance and academic integrity offices—especially as students increasingly treat take-home assessments as an AI-assisted production environment rather than a learning signal.