{"resourceId":"aaup-aft-ai-tech-audit-guide-202609","versions":[{"version":"external-3eb6d913cd175510d847cca5465c0e78818f13b1b47f588658c21eeb5d8d83e8","resource":{"id":"aaup-aft-ai-tech-audit-guide-202609","title":"AAUP/AFT AI/Tech Audit Guide equips campus workers to independently audit GenAI and edtech contracts","organization":"American Association of University Professors (with AFT Higher Ed)","sector":"Higher education labor / shared governance","geography":"United States (national faculty/staff association guidance)","publishedAt":"September 2026 (toolkit webinar September 9, 2026; PDF under /2026-09/)","publicationDate":"2026-09-09","eventDate":"2026-09-09","sourceName":"AAUP ad hoc Committee on AI and Academic Professions","sourceLabel":"AI/Tech Audit Guide PDF in AAUP Toolkit for Activism on AI","sourceUrl":"https://www.aaup.org/sites/default/files/2026-09/AAUP-AFT-Higher-Ed-AI-Tech-Audit-Guide.pdf","evidenceClass":"public-sector-association","outcomeClass":"cautionary","topics":["governance-procurement","accessibility-workforce","operating-model","data-security","knowledge-work"],"finding":"AAUP's September 2026 AI/Tech Audit Guide (ad hoc Committee on AI and Academic Professions; introduced in the September 9 toolkit webinar) argues that ongoing independent audits of campus technologies—from LMS platforms to generative AI partnerships—minimize harms before they occur. It instructs members to obtain vendor contracts (via FOIA/open records where needed), triangulate vendor claims with worker experience across job types, and complete structured worksheets on data privacy (including FERPA/COPPA/HIPAA handling), opt-in/out control, intellectual property, technical methods, institutional resource costs (compute, water, electricity), social impacts (accessibility, surveillance, differential effects by job category), and economic/labor impacts (cost, affiliate relationships, unseen labor, precarity, deskilling, shared-governance gaps). The guide is designed to surface disparities between administrator/vendor claims and grounded outcomes.","sledRelevance":"Direct campus-operations labor and governance constraint: procurement and GenAI rollouts face organized demands for independent contract/labor/privacy audits—not only CIO architecture checklists. Why now: September 2026 toolkit materials were not archived; prior campus-operations run could not open the related AAUP topical report (timeout).","evidence":"Official AAUP PDF practice guide plus toolkit/webinar context (slides dated September 9, 2026). Prescriptive methodology without publishing results of a completed campus audit in this document.","architectureImplications":"Expect audit requests for logging, training-data use, resource-cost disclosure, and human-oversight paths; design GenAI platforms so contract terms and technical controls are inspectable by non-vendor parties.","governanceImplications":"Shared-governance and bargaining actors are being equipped to challenge top-down tech adoption; institutions should plan participatory review of AI/edtech procurement rather than post-hoc justification.","securityPrivacyImplications":"Guide highlights school-official exceptions and weak enforcement language in edtech contracts—campuses should clarify data ownership, access, and FERPA/HIPAA boundaries in AI agreements.","caveats":"Advocacy toolkit, not an empirical multi-campus study. Effectiveness depends on local open-records rights and bargaining power. International transfer requires local labour-law mapping.","streamIds":["campus-operations"],"roles":{"sales":"Faculty/staff unions are mobilizing independent audits of AI and edtech deals. Buyers who cannot show contracts, privacy terms, labor-impact analysis, and resource-cost disclosures will face delay. Position engagements as audit-ready procurement packages—not booster demos. Unsupported: claiming the guide endorses any vendor product.","engineering":"Engineer for inspectability: exportable logs, clear data-flow diagrams, documented human oversight, and answers to privacy/IP/labor worksheet questions. Prerequisites include contract repositories and data-classification maps. Constraints: guide does not specify technical standards. Proof of value is producing a filled audit worksheet for one production AI use case.","delivery":"Delivery includes standing up a cross-functional audit response team (procurement, counsel, IT, HR, faculty senate/union liaisons), FOIA-ready contract packs, and labor-impact reviews before go-live. Ownership should be shared between CIO/procurement and governance bodies. Acceptance criteria: completed privacy/labor/cost worksheets and published human-authority rules for consequential decisions. Risks: adversarial FOIA cycles and stalled deployments if documentation is missing."},"roleDetails":{"sales":{"customerProblem":"Administrations face organized demands for independent audits of AI/edtech costs, contracts, privacy, and labor impacts before or during adoption.","stakeholders":["Faculty senate / AAUP chapter","Unions / staff associations","CIO and procurement","General counsel","HR / academic affairs"],"discoveryQuestions":["Can the institution produce AI/edtech contracts and correspondence on request?","Has a labor-impact assessment been done for the proposed tool?","Who can opt out, and where is that documented?","Are compute/energy/water cost questions part of procurement?"],"valueHypothesis":"Audit-ready documentation and participatory review can reduce later bargaining conflicts and FOIA surprises—without guaranteeing smoother ratification in this source.","potentialEngagement":"Four-week AI procurement audit-readiness sprint producing filled AAUP-style worksheets for one priority platform.","unsupportedClaims":["Union endorsement of a specific AI product","Guaranteed avoidance of labor disputes","Measured productivity benefits from auditing alone"]},"engineering":{"fit":"Institutional GenAI/edtech systems subject to shared-governance and labor scrutiny.","architectureIntegration":"Systems should expose data flows, retention, training-use flags, access controls, and human-in-the-loop points that map to audit worksheet fields.","prerequisites":["Searchable contract repository","Data classification and FERPA/HIPAA mappings","Named human decision owners for consequential outputs"],"constraints":["Guide is procedural, not a technical standard","Public-records scope varies by state","Vendor NDAs may conflict with transparency demands"],"security":"Clarify school-official designations, subprocessors, and enforcement remedies in contracts; avoid opaque model training on campus data.","proofOfValue":["Completed audit spreadsheet for one live AI use case","Documented opt-out and human-authority paths","Resource-cost estimate (compute/energy) attached to the buy"]},"delivery":{"work":"Assemble contract packs, run participatory audit sessions, document labor/privacy findings, and feed results into go/no-go decisions.","dependencies":["Open-records/FOIA process","Vendor cooperation","Union/senate calendars","Security and privacy offices"],"ownership":"Joint ownership: procurement/CIO for documentation; governance/labor bodies for independent review findings.","skillsAdoption":"Train chapter/senate volunteers and IT staff on worksheet methods and FOIA request hygiene.","governanceCheckpoints":["Pre-procurement audit kickoff","Mid-implementation labor/privacy review","Annual re-audit when vendor capabilities change"],"acceptanceCriteria":["Worksheet domains completed with cited sources","Published summary of unsupported vendor claims found","Human final-authority statement for employment/academic decisions"],"risks":["Polarized campus conflict without facilitation","Incomplete vendor disclosures","Deployment delays that push shadow IT"]}},"retrievedAt":"2026-09-19T03:09:00Z","enrichedAt":"2026-09-19T03:13:40Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Worksheets explicitly ask about accessibility, differential harms by job category, unseen labor, precarity, and deskilling—centering workforce impacts that platform rollouts often omit.","procurementImplications":"Treat FOIA-ready contracts, privacy clauses, cost disclosures, and labor-impact questions as first-class procurement artifacts for AI/edtech, not post-signature afterthoughts.","operatingModelImplications":"Shifts AI adoption from admin-led enablement toward contested, auditable shared-governance processes with member-led verification.","sourceVerification":{"openedUrl":"https://www.aaup.org/sites/default/files/2026-09/AAUP-AFT-Higher-Ed-AI-Tech-Audit-Guide.pdf","referenceExcerpt":"Ongoing independent audits of technology minimize harms before they happen.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}