{"resourceId":"uniosun-institutional-ai-policy-tiers-detector-limits-20261107","versions":[{"version":"external-d7dceac8da99a68c3841db7b63297fb37c611f6bbb4ecf73a1ee8d01907d0396","resource":{"id":"uniosun-institutional-ai-policy-tiers-detector-limits-20261107","title":"Osun State University adopts institution-wide AI policy: mandatory disclosure, per-course three-tier AI designations, exam ban, and detector flags never sole proof of misconduct","organization":"Osun State University (UNIOSUN)","sector":"Higher education institutional AI policy and academic integrity","geography":"Nigeria","publishedAt":"UNIOSUN website, policy effective November 2026 (page modified 7 November 2026 UTC)","publicationDate":"2026-11-07","eventDate":null,"sourceName":"Osun State University","sourceLabel":"University policy page","sourceUrl":"https://uniosun.edu.ng/institutional-artificial-intelligence-ai-policy/","evidenceClass":"standards-guidance","outcomeClass":"emerging","topics":["governance-procurement","data-security","knowledge-work"],"finding":"UNIOSUN’s institutional AI policy (effective November 2026) covers all students, staff and visiting researchers. Students must declare when and how generative AI was used; submitting unedited AI-generated text, code or data as original work is plagiarism. Lecturers must state on every course outline one of three tiers: prohibited, assisted/permitted (brainstorming, outlining, grammar with mandatory citation) or fully integrated (with a reflection appendix). Generative AI and smart devices are banned in physical and remote-proctored exams unless the Senate authorises specific computer-based tests. Staff are urged toward authentic assessment (viva voce, localized case studies, reflective journals, in-class problem solving). The policy recognises detector limits: flags are never sole proof of misconduct and require holistic review and student dialogue. A multidisciplinary committee reviews guidelines annually.","sledRelevance":"An international institutional policy converging on the same controls U.S. campuses are adopting—tiered course designations, disclosure and human review over detector-only findings—useful as a policy-design benchmark, with transferability limits.","evidence":"Primary policy document; describes rules, not implementation or outcomes.","architectureImplications":"Implies course-outline metadata for AI tiers and proctoring controls for computer-based exams.","governanceImplications":"Course-level tier declarations, annual committee review, and routing violations through standard disciplinary committees.","securityPrivacyImplications":"Research use must comply with ethics guidelines on data privacy, consent and algorithmic bias; encourages local datasets.","caveats":"No exact effective date; enforcement capacity unknown; Nigerian context and resource constraints limit transfer to U.S. institutions.","streamIds":["student-success"],"roles":{"sales":"Sales framing: universities drafting AI policies want a coherent model that faculty can apply course by course. UNIOSUN’s policy is a compact benchmark—disclosure, three tiers, exam limits and an explicit rule that detector flags are never sole proof. A bounded engagement is policy benchmarking and tier-implementation support; the document offers no evidence that the policy reduces misconduct or improves learning.","engineering":"Engineering framing: fit is metadata and workflow rather than AI models—recording a tier on each course outline, capturing disclosure statements with submissions, and locking down authorised computer-based tests. Any detector in use must be configured as advisory only, feeding a human review queue. Validation should confirm tier data is visible to students at submission time and that review workflows document student dialogue.","delivery":"Delivery framing: an academic planning office and the new multidisciplinary committee own rollout, with lecturers declaring tiers. Delivery requires templates, faculty training on tier selection and authentic assessment, and a review cadence. Proposed acceptance measures include tier coverage across course outlines, disclosure completion and audit of misconduct cases to confirm none rest on detector flags alone."},"roleDetails":{"sales":{"customerProblem":"Institutions lack consistent, enforceable AI rules that faculty can apply per course without over-relying on detectors.","stakeholders":["Academic planning","Senate","Faculty deans","Library services","Student affairs"],"discoveryQuestions":["Do course outlines carry an AI-use designation?","How are detector flags handled in misconduct cases?","Who reviews AI policy annually?"],"valueHypothesis":"Interpretation: tiered declarations with human-review rules may improve clarity and fairness; untested in this source.","potentialEngagement":"Policy benchmark and tier-template rollout support for one faculty.","unsupportedClaims":["Reduced misconduct","Improved learning outcomes","Direct transfer to U.S. regulatory context"]},"engineering":{"fit":"Course-outline metadata, disclosure capture and advisory-only detector configuration.","architectureIntegration":"LMS/course catalogue field for AI tier; disclosure field on submissions; exam lockdown for authorised CBTs.","prerequisites":["Approved policy","Course outline templates","Disciplinary workflow"],"constraints":["Nigerian infrastructure and resource context","No exact effective date","Enforcement capacity unknown"],"security":"Keep disclosures and misconduct records under institutional data protection; avoid sending student work to unvetted detectors.","proofOfValue":["Audit tier coverage on outlines","Sample misconduct cases for human-review evidence","Student survey on rule clarity"]},"delivery":{"work":"Publish templates, train lecturers, stand up committee, monitor cases and update annually.","dependencies":["Governance committee","Lecturers","Disciplinary committees","Library services"],"ownership":"Academic planning office with the University AI Governance Committee","skillsAdoption":"Faculty training on tier choice, authentic assessment and AI literacy embedding.","governanceCheckpoints":["Senate approval of exam exceptions","Annual committee review","Misconduct case audit"],"acceptanceCriteria":["Proposed: AI tier stated on 100% of pilot course outlines","Proposed: zero misconduct findings resting solely on detector flags","Proposed: annual review report published"],"risks":["Uneven faculty compliance","Detector over-reliance in practice","Exam ban hard to enforce remotely"]}},"retrievedAt":"2026-11-08T04:19:10Z","enrichedAt":"2026-11-08T04:22:57Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Encourages embedding prompt literacy and critical evaluation in courses; staff will need training to assign tiers and redesign assessment.","procurementImplications":"Detector limits language cautions against detector-led procurement; proctoring tools may be needed for authorised computer-based tests.","operatingModelImplications":"Faculty must declare tiers on every course outline; governance committee spans computing, academic planning, library and student affairs.","peerReview":"unknown","sourceVerification":{"openedUrl":"https://uniosun.edu.ng/institutional-artificial-intelligence-ai-policy/","referenceExcerpt":"Detection flags will never be used as sole proof of misconduct; they must be paired with holistic review and student dialogue.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}