{"resourceId":"iagst-university-ai-governance-breakpoints-20260828","versions":[{"version":"external-71a3c0f10c6799519aa6f1b73610087d291f2468da219b988a09aa159dabb753","resource":{"id":"iagst-university-ai-governance-breakpoints-20260828","title":"IAGST: 55/75 university AI governance cases indeterminate; pathways collapse at autonomous substitution","organization":"Charles Darwin University (Biranchi Poudyal)","sector":"Higher education AI governance / research administration","geography":"Australia (five Western Australian universities demonstration); transfer limits to U.S. SLED","publishedAt":"August 28, 2026 (arXiv:2608.28925)","publicationDate":"2026-08-28","eventDate":null,"sourceName":"arXiv","sourceLabel":"Academic preprint — institutional AI governance stress-testing method and WA demonstration","sourceUrl":"https://arxiv.org/pdf/2608.28925","evidenceClass":"academic-research","outcomeClass":"cautionary","topics":["governance-procurement","operating-model","developers-agents","knowledge-work"],"finding":"Institutional AI Governance Stress Testing (IAGST) prospectively tests whether university documents yield accountable responses as AI agency escalates. Using 133 substantive public documents from five Western Australian universities and 15 quality-screened scenarios (75 university-scenario encounters), six cases (8.0%) were resolved, 14 (18.7%) were resolved through structured discretion, and 55 (73.3%) were indeterminate. Governed pathways fell from 16 of 25 augmentation cases to four of 25 delegation cases and none at autonomous substitution. All 50 authority-gap cases named a role only at a generic level without sufficient decision criteria or process. The author positions IAGST as a reproducible diagnostic for policy learning, not a ranking or implementation measure.","sledRelevance":"Why now: Still-unarchived research-administration governance evidence newly relevant after the 2026-09-21 physical-lab and federated-HPC agent edition—pairs capacity and agent tooling with documentary breakpoint risk as campuses adopt research AI agents.","evidence":"133 documents; 15 scenarios; 75 encounters; 6 resolved / 14 structured discretion / 55 indeterminate (73.3%); 16/25→4/25→0/25 governed pathway escalation; 50 authority-gap generic-role cases.","architectureImplications":"Agentic research tooling deployments should be accompanied by scenario libraries that escalate from assistants to cross-system agents so governance gaps are tested before production autonomy.","governanceImplications":"Move beyond AI principle inventories to connected authority, procedure, safeguard, and review pathways; generic role naming is insufficient under delegation/autonomy stress.","securityPrivacyImplications":"Indeterminate pathways for cross-system agents imply unclear ownership for data access, monitoring, and incident response—raise before enabling persistent research agents.","caveats":"August preprint with why-now; WA public documents only; documentary preparedness ≠ lived practice; Australian HE transfer limits.","streamIds":["research"],"roles":{"sales":"Sales framing: Research administration leaders writing AI principles need accountable decision pathways—not thicker policy PDFs. IAGST's demonstration supports a cautionary conversation: across 75 university-scenario encounters on 133 Western Australian public documents, 55 (73.3%) were indeterminate, and governed pathways fell from 16/25 augmentation cases to none at autonomous substitution, with 50 authority-gap cases naming only generic roles. Discovery should map which research-AI agency levels the campus already faces and where procedures/safeguards break. Unsupported: ranking U.S. campuses from WA documents, that indeterminate equals no informal practice, or that any vendor tool fixes governance breakpoints.","engineering":"Engineering framing: Fit is a prospective documentary stress-test method (frozen corpus, capability escalation, six-dimensional non-compensatory response chain, breakpoint signatures)—not an IT product install. Integration points are policy repositories, research-integrity procedures, and scenario libraries for agentic research tools. Prerequisites include a frozen public/internal document set, agreed escalation levels (augmentation→delegation→autonomous substitution), and scoring anchors. Constraints: WA five-university demonstration; documentary preparedness ≠ implementation. Security/privacy scenarios must be included when agents cross systems. Proof of value is movement of breakpoints after reform—not a one-time maturity score.","delivery":"Delivery framing: Run a bounded IAGST-style exercise on the campus research-AI document set with 10–15 scenarios spanning literature assistants to cross-system agents; record breakpoints and owners. Dependencies: research integrity, general counsel, CIO/research computing, and faculty senate. Ownership: research administration owns remediation roadmap; units own local procedures. Skills adoption trains policy owners to write decision criteria, not only principles. Governance checkpoints: freeze corpus, dual-score cases, publish remediation owners. Acceptance: documented breakpoints with owners and a retest date; risks include theater without procedure updates and over-transfer from Australian HE context."},"roleDetails":{"sales":{"customerProblem":"Universities issue AI principles faster than they build decision pathways for unfamiliar AI agency in research settings.","stakeholders":["VPR / research integrity","General counsel","CIO / research computing","Faculty governance"],"discoveryQuestions":["Which AI agency levels are in use now?","Where do documents name only generic roles?","What review/remedy paths exist?","When will breakpoints be retested?"],"valueHypothesis":"Documentary stress-testing can locate where authority/procedure/safeguard chains break (55/75 indeterminate; 0/25 autonomous substitution governed in the demo) if results drive remediation—without claiming lived-practice measurement.","potentialEngagement":"Bounded research-AI governance stress-test with scenario pack and remediation owners.","unsupportedClaims":["WA results rank U.S. campuses","Indeterminate means no informal practice","Tool purchase closes breakpoints"]},"engineering":{"fit":"Prospective documentary method combining frozen corpus, capability escalation, and non-compensatory governance-chain scoring.","architectureIntegration":"Map policy corpus and research-AI tool inventory to scenario library; store breakpoint signatures and remediation tickets in governance systems of record.","prerequisites":["Frozen document corpus","Escalation-level definitions","Scoring anchors","Retest schedule"],"constraints":["WA public-document demo","Not an implementation metric","Transfer limits to U.S. SLED"],"security":"Include cross-system agent and data-access scenarios so privacy/security breakpoints surface with research-integrity ones.","proofOfValue":["Breakpoint register with owners","Retest showing moved breakpoints","Reduction in generic-only authority cases"]},"delivery":{"work":"Freeze corpus, run screened scenarios, score dimensions, assign remediation owners, and schedule retest.","dependencies":["Research integrity","Counsel","CIO/research computing","Faculty governance"],"ownership":"Research administration owns the remediation roadmap; units own local procedure updates.","skillsAdoption":"Train policy authors to specify decision criteria, processes, safeguards, and reviews.","governanceCheckpoints":["Corpus freeze","Dual coding / validation","Public remediation owners","Retest date"],"acceptanceCriteria":["Breakpoint register complete","Owners assigned","Retest calendar set"],"risks":["Policy theater without procedure changes","Over-transfer from WA context","Ignoring informal workarounds"]}},"retrievedAt":"2026-09-22T22:17:16-05:00","enrichedAt":"2026-09-22T22:17:16-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Clear decision pathways reduce ad-hoc burden on individual faculty and RSEs confronting novel agent behaviors; without them, risk concentrates on the least powerful actors.","procurementImplications":"Require vendors of research-AI agents to map features to campus decision rights and escalation procedures—not only acceptable-use clauses.","operatingModelImplications":"Creates a recurring foresight-to-remediation loop owned by research administration, with units updating procedures where breakpoints appear.","sourceVerification":{"openedUrl":"https://arxiv.org/pdf/2608.28925","referenceExcerpt":"Six cases were resolved, 14 were resolved through structured discretion, and 55 were indeterminate","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}