{"resourceId":"daily-northwestern-mccormick-professors-in-person-exams-after-brown-ai-cheating-20261020","versions":[{"version":"external-b6b4feace80a4c58e62d50b84a25e4666f291e76dd17ce24ae0608d580281e97","resource":{"id":"daily-northwestern-mccormick-professors-in-person-exams-after-brown-ai-cheating-20261020","title":"Northwestern professors shift to in-person and paper exams, version-history checks and AI disclosure after Brown's take-home exam AI cheating episode","organization":"Northwestern University (McCormick School of Engineering and Weinberg College); The Daily Northwestern","sector":"Higher education assessment and academic integrity","geography":"Evanston, Illinois, United States","publishedAt":"The Daily Northwestern, 20 October 2026","publicationDate":"2026-10-20","eventDate":null,"sourceName":"The Daily Northwestern","sourceLabel":"Student newspaper reporting on faculty assessment responses to AI","sourceUrl":"https://dailynorthwestern.com/2026/10/20/campus/mccormick-professors-adapt-after-brown-ai-cheating-scandal/","evidenceClass":"independent-reporting","outcomeClass":"cautionary","topics":["knowledge-work","governance-procurement","operating-model"],"finding":"The 20 October article relays, from Fortune, that a Brown economics professor's spring take-home midterm averaged 96% against 65% to 80% in earlier years, with 40 of 86 students scoring perfect marks and most answers matching ChatGPT output; when the final moved in person, nearly a third of the class dropped or did not sit it and the average fell to 48%. At Northwestern, McCormick professor Zach Wood-Doughty now feels compelled to use in-person or paper exams in 300- and 400-level classes. A student reports a professor reviewing Google Docs version histories. Professor Edith Elkind asks students to disclose AI use without grade penalty and says completion grading reduces homework cheating.","sledRelevance":"Shows how a widely reported unsupervised-exam failure is changing assessment practice at another research university, alongside lighter-touch disclosure approaches. Relevant to U.S. course policy and integrity process design.","evidence":"Student journalism with named faculty and student sources; Brown statistics are second-hand from Fortune. Instructor observations, not studies.","architectureImplications":"Interpretation: version-history review and disclosure workflows depend on document platforms (Google Docs) and LMS settings rather than detectors.","governanceImplications":"Practices vary by instructor: in-person exams, version-history review, and penalty-free disclosure. No school-wide policy is reported.","securityPrivacyImplications":"Reviewing document version histories exposes students' drafting process; consent and retention expectations are not described.","caveats":"Anecdotal and instructor-level; the Brown figures come from another outlet. Completion grading reducing cheating is one professor's observation.","streamIds":["student-success"],"roles":{"sales":"Sales framing: the Brown episode, as relayed here, shows how take-home assessments can fail under AI, and Northwestern faculty are reacting course by course. Use it to open discovery on assessment policy consistency, proctoring capacity and disclosure processes. Offer bounded help mapping assessment risk across gateway courses. Do not claim that in-person exams or disclosure improve learning or integrity: the article reports instructor choices and second-hand figures, not evaluated outcomes.","engineering":"Engineering framing: the controls described are low-tech: in-person or paper exams, Google Docs version-history review and AI-use disclosure. Engineering support is about configuring document and LMS tools to capture drafting history and disclosure statements consistently, and about privacy settings for that evidence. Validate that version-history evidence is reliable and that students are told when it will be reviewed, rather than adding AI detectors.","delivery":"Delivery framing: instructors are changing assessment individually, so delivery work is coordination. Inventory which courses use take-home exams, agree a small set of approved controls, plan proctoring and accommodation capacity for in-person exams, and publish disclosure expectations. Track grade distributions, withdrawal rates and integrity cases against prior terms to see whether changes help, since the reporting offers no measured effects."},"roleDetails":{"sales":{"customerProblem":"Take-home assessments may no longer show what students can do, and faculty responses are inconsistent.","stakeholders":["Associate deans for undergraduate education","Academic integrity office","Testing and accommodations centre","Faculty in high-enrolment courses","Student government"],"discoveryQuestions":["Which courses rely on unsupervised take-home exams?","What proctoring and accommodation capacity exists?","Is there a standard AI disclosure format?","How are integrity cases trending?"],"valueHypothesis":"Interpretation: a coordinated assessment-risk review could make controls more consistent and fair than course-by-course changes.","potentialEngagement":"A bounded assessment-risk inventory and policy alignment exercise for one school or college.","unsupportedClaims":["Reduced cheating from in-person exams","Learning gains","Reliability of version-history evidence","Brown figures verified at source"]},"engineering":{"fit":"Fits institutions using shared document platforms and LMS tools that can record drafting history and disclosures.","architectureIntegration":"Assessment evidence from Google Docs version histories and AI-use disclosure statements, plus in-person exam logistics.","prerequisites":["Document platform settings","LMS assignment configuration","Proctoring space and staff","Disclosure template"],"constraints":["Instructor-level practices only","Second-hand statistics","Privacy expectations for drafting data","Accommodation requirements for in-person exams"],"security":"Limit access to version histories to the instructor, define retention, and tell students in advance when drafting history will be reviewed.","proofOfValue":["Grade distribution comparison with prior terms","Withdrawal rate tracking after format change","Integrity case volume and outcomes","Student survey on fairness"]},"delivery":{"work":"Coordinate assessment controls across courses and plan proctoring and disclosure processes.","dependencies":["Faculty participation","Testing centre capacity","Integrity office","Document and LMS platforms"],"ownership":"Associate dean for undergraduate education with the academic integrity office.","skillsAdoption":"Faculty guidance on approved controls and student guidance on disclosure and drafting evidence.","governanceCheckpoints":["School-level assessment policy review","Accommodation impact review","Privacy review of drafting evidence","End-of-term review of outcomes"],"acceptanceCriteria":["Proposed: every gateway course lists its AI assessment control in the syllabus","Proposed: accommodation requests for in-person exams met in full","Proposed: withdrawal rates monitored and reviewed each term","Proposed: disclosure format used across participating courses"],"risks":["Higher withdrawals after format change","Inconsistent rules confusing students","Privacy complaints about version histories","Proctoring overload"]}},"retrievedAt":"2026-10-21T03:22:21Z","enrichedAt":"2026-10-21T03:27:54Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"In-person and paper exams add proctoring and accommodation workload; this is not discussed in the article.","procurementImplications":"Limited: favours assessment design and existing tools over detector purchases.","operatingModelImplications":"Assessment changes are being made course by course, which risks inconsistent student experience without coordination.","sourceVerification":{"openedUrl":"https://dailynorthwestern.com/2026/10/20/campus/mccormick-professors-adapt-after-brown-ai-cheating-scandal/","referenceExcerpt":"Following the scandal, McCormick Prof. Zach Wood-Doughty said he feels compelled to have in-person or on-paper exams in his 300- and 400-level classes.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}