The department asks the StudyBuddy team to add a feature: "Automatically check every Q&A post and lab report with an AI-text detector. Posts scoring above 80% 'AI-generated' are flagged, and the instructor receives a weekly list of flagged students for possible academic-misconduct proceedings." A vendor offers a detector claiming "99% accuracy".
Write an ethical and legal analysis (about one page):
- Stakeholders and how each is affected (include those with no voice in the decision).
- Technical concerns: what "99% accuracy" hides (base rates, false positives, non-native writers — use the Stanford 2023 finding), and what evidence you would demand.
- Code of Ethics: which ACM/IEEE principles are involved and how.
- Legal view: privacy (Chapter 11), EU AI Act classification of such a system, transparency duties.
- Your recommendation to the department: build / do not build / build with conditions — with concrete conditions or alternatives.