{"resourceId":"coventry-university-group-ai-adoption-lab-baselines-20261005","versions":[{"version":"external-509c29dbeaf768b8fb0b0273255044dcf929f00f41bdd568f6b7f768003dcf7b","resource":{"id":"coventry-university-group-ai-adoption-lab-baselines-20261005","title":"Coventry University Group: AI Adoption Lab requires outcome, baseline and owner for 13 live use cases; about 2,900 staff in Essential AI learning","organization":"Coventry University Group","sector":"Public-funded university group / institutional AI operating model","geography":"United Kingdom","publishedAt":"Monday 05 October 2026","publicationDate":"2026-10-05","eventDate":null,"sourceName":"Coventry University news (opinion)","sourceLabel":"Institutional first-party opinion by the Group's AI lead (self-reported)","sourceUrl":"https://www.coventry.ac.uk/news/2026/opinion-world-ai-week-2026/","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["operating-model","accessibility-workforce","governance-procurement","knowledge-work"],"finding":"Coventry's AI lead, Prof. Elena Gaura, describes a 'govern centrally and deploy locally' model: the Group sets policy, controls, approved platforms and escalation routes; teams own low-risk experimentation; uses involving sensitive data, decisions about people, procurement, integration or wider deployment get formal technical and ethical risk assurance. About 2,900 colleagues have engaged with Essential AI learning since July 2026, more than 2,000 attended AI tools training since Christmas 2025, and colleagues from half of organisational units are on a Level 4 AI and Automation Practitioner apprenticeship. The AI Adoption Lab supports 13 live use cases across 10 units; each defines outcome, baseline and owner, and tracks quality, staff time and financial effects, with embedded doctoral and postdoctoral researchers. No use-case results are reported.","sledRelevance":"Operating-model reference for campus AI: measurement and ownership are required before scaling, risk assurance is tiered by data and decision type, and workforce capability is built at scale. Useful as a design pattern, not evidence of benefit.","evidence":"Opened Coventry University's Oct. 5, 2026 opinion article with governance tiers, training participation counts and AI Adoption Lab method.","architectureImplications":"The article identifies data that cannot move across siloed systems as the main barrier; integration is one of the triggers for formal assurance. No platform details are given.","governanceImplications":"Tiered assurance: local ownership for low-risk uses, formal technical and ethical assurance for sensitive data, decisions about people, procurement and integration. Each use case records approvals and measures for reuse.","securityPrivacyImplications":"Sensitive-data uses are routed to formal assurance; the article gives no detail on controls, incidents or data-protection review.","caveats":"Self-reported opinion that also promotes Coventry's services to businesses; participation counts are activity, not outcomes; no use-case names, results, baselines or costs are published. UK context limits transferability to U.S. governance structures.","streamIds":["campus-operations"],"roles":{"sales":"Sales framing: the useful message is measure-first adoption. Coventry requires an outcome, baseline and owner for each of 13 use cases before choosing where AI fits, and tiers risk assurance by data and decision type. Use this to propose a bounded adoption-lab design with institutions stuck at pilot stage. Avoid presenting it as proof of savings: the article reports participation counts and method, not results.","engineering":"Engineering framing: integration, not model capability, is named as the barrier. Fit is institutions whose administrative processes cross siloed systems. Engineering should map the end-to-end process and data flows for each use case, flag integration and sensitive-data triggers for formal assurance, and record reusable components. Proof of value should compare against the declared baseline for quality and staff time, which Coventry says it tracks but has not published.","delivery":"Delivery framing: each use case needs an owner and a reusable record. Coventry's approach pairs local teams with embedded researchers and documents skills, data, approvals and measures so others can adapt them. Delivery should establish an intake template with baseline fields, a tiered assurance route and a reuse library. Risks include training activity substituting for outcomes, researcher dependency, and stalled scaling when data or procurement barriers sit outside the team's remit."},"roleDetails":{"sales":{"customerProblem":"Institutions run AI pilots that work technically but stall because no one owns the end-to-end change, data cannot move and value is not measured.","stakeholders":["Provost or deputy vice-chancellor for operations","CIO and data lead","HR and learning and development","Process owners in professional services","Risk and ethics committee"],"discoveryQuestions":["Do current AI pilots have a declared baseline, outcome and owner?","Which uses trigger formal risk assurance today?","How are successful use cases documented for reuse?","What share of staff has completed applied AI training?"],"valueHypothesis":"Interpretation: a measure-first adoption lab could convert stalled pilots into evaluated services; Coventry has not published results that quantify this.","potentialEngagement":"A bounded adoption-lab setup: intake template with baselines, tiered assurance route and evaluation of three to five administrative use cases.","unsupportedClaims":["No use-case results or savings are reported.","Training counts do not show productivity change.","Transfer to U.S. governance structures is not shown."]},"engineering":{"fit":"Administrative processes spanning several teams and systems where handovers and data silos limit AI value.","architectureIntegration":"End-to-end process mapping, data-flow integration across team systems, approved platforms, and documented integrations reusable by other teams.","prerequisites":["Process maps with handover points","Accessible, consistently described data","Approved AI platforms","Baseline measures for time and quality"],"constraints":["Incompatible data formats across teams","Security controls limiting data access","UK data-protection requirements"],"security":"Route sensitive-data and people-decision uses to formal technical and ethical assurance before deployment.","proofOfValue":["Compare each use case to its declared baseline for staff time and quality","Record integration effort per use case","Track reuse of documented components by other teams"]},"delivery":{"work":"Stand up an adoption lab with intake, baselines, owners, tiered assurance and a reuse library for administrative AI use cases.","dependencies":["Process owners","Data and integration teams","Researchers or analysts for evaluation","Risk assurance function"],"ownership":"A central AI lead with each use case owned by its operating unit.","skillsAdoption":"Staff move from awareness training to applied practice; champions and apprenticeships spread capability beyond the central team.","governanceCheckpoints":["Baseline and owner recorded before build","Formal assurance for sensitive data, people decisions, procurement or integration","Measured value review before scaling"],"acceptanceCriteria":["Proposed: 100% of use cases with recorded baseline, outcome and owner","Proposed: published results for each use case at scale-up decision","Proposed: at least one reuse of each successful use-case pattern"],"risks":["Training activity mistaken for outcomes","Dependence on embedded researchers","Scaling blocked by data or procurement barriers"]}},"retrievedAt":"2026-10-05T21:53:22-05:00","enrichedAt":"2026-10-05T21:53:22-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Large-scale staff development (about 2,900 in Essential AI learning, 2,000+ in tools training) and an apprenticeship route; people remain accountable for purpose, exceptions and consequences.","procurementImplications":"Procurement is one of the triggers for formal assurance; use cases that expose procurement disconnects are escalated to leadership.","operatingModelImplications":"Central governance with local deployment and documented reuse; capability is distributed through champions rather than held in a single central team.","sourceVerification":{"openedUrl":"https://www.coventry.ac.uk/news/2026/opinion-world-ai-week-2026/","referenceExcerpt":"Each use case team defines the outcome, baseline and owner before deciding where AI could improve the process.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}