{"resourceId":"surrey-lumi-module-grounded-ai-study-assistant-surreylearn-20261019","versions":[{"version":"external-ac017d642f786e00ac90cb0b88a783b7f72f3ead716fc7aa12f3340fc443b854","resource":{"id":"surrey-lumi-module-grounded-ai-study-assistant-surreylearn-20261019","title":"University of Surrey launches Lumi in SurreyLearn: a module-grounded AI study assistant that cannot complete assessed work or search the internet","organization":"University of Surrey","sector":"Higher education AI study support / course-grounded tutoring in the LMS","geography":"Guildford, United Kingdom","publishedAt":"MySurrey student website news, Monday 19 October 2026","publicationDate":"2026-10-19","eventDate":null,"sourceName":"University of Surrey (MySurrey student website)","sourceLabel":"University student-news announcement of an LMS-embedded AI study assistant","sourceUrl":"https://my.surrey.ac.uk/news/meet-lumi-your-ai-powered-study-assistant","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["knowledge-work","data-security","governance-procurement","operating-model"],"finding":"Surrey announced on 19 October that Lumi is now available in SurreyLearn. As described, it explains concepts, tests understanding, generates summaries, quizzes, flashcards and practice questions, finds module resources, reviews module structure, key dates and deadlines, and answers questions about university services and processes. It bases answers only on module resources rather than the wider internet. The university states that Lumi cannot complete assessed work, replace lecturers, search external websites, give feedback on uploaded documents or replace academic judgement, and warns it can make mistakes. Surrey's AI tools page says Lumi appears on SurreyLearn pages and links answers to course materials. No usage or outcome data are given.","sledRelevance":"A same-day example of the course-grounded, LMS-embedded study assistant model, with explicit scope limits (no assessed work, no internet, no document feedback). U.S. institutions using Brightspace or similar LMS AI features face the same scoping and evaluation choices.","evidence":"University announcement of feature scope and limits. No usage, accuracy, satisfaction or learning-outcome evidence.","architectureImplications":"Interpretation: retrieval is restricted to module content inside the LMS, so answer quality depends on how complete and current each module's materials are. Module dates and deadlines become a data-quality dependency.","governanceImplications":"Scope limits are published to students (no assessed work, no feedback on uploads), which sets expectations for academic-integrity policy. Who owns accuracy per module is not stated.","securityPrivacyImplications":"Grounding in module content and no internet search narrow data exposure, but retention of student prompts and vendor processing terms are not described.","caveats":"Announcement only. The Surrey page does not name a vendor; D2L's customer case study identifies SurreyLearn as Surrey's Brightspace instance and D2L markets a product called D2L Lumi, so Lumi is very likely the platform vendor's feature (interpretation, not stated by Surrey). Benefits are intended functions, not measured outcomes. UK regulatory context.","streamIds":["student-success"],"roles":{"sales":"Sales framing: Surrey shows that universities now want AI study help that stays inside the LMS and course materials, with published limits on assessed work. Use it to open discovery on LMS platform AI features, module content quality and academic-integrity expectations. Offer bounded help with configuration, content readiness and evaluation design. Do not cite it as evidence of better learning, engagement or retention: the announcement reports no usage or outcome data and only describes intended functions.","engineering":"Engineering framing: Lumi is a retrieval-grounded assistant restricted to module resources, with no internet access and no document upload feedback. The engineering questions are about retrieval quality across uneven module content, citation back to source materials, handling of dates and deadlines, and logging. Validate answer accuracy per module, check that assessment items cannot be surfaced as answers, and confirm prompt retention and processing terms with the platform provider before enabling the feature widely.","delivery":"Delivery framing: launching an LMS-embedded assistant moves responsibility for answer quality onto module teams. Plan a staged rollout by module, with content readiness checks, a student-facing route to report wrong answers, and a named owner for corrections. Track usage, error reports and student feedback against a baseline. Pair the launch with clear guidance on permitted use so students understand that the assistant supports study and does not replace teaching."},"roleDetails":{"sales":{"customerProblem":"Students need help with course content outside teaching hours, but institutions fear open AI tools undermine integrity and give unreliable answers.","stakeholders":["Pro vice-chancellor or vice provost for education","Digital learning and LMS team","Academic integrity lead","Module leaders","Students' union or student representatives"],"discoveryQuestions":["Which LMS AI features are already licensed but not enabled?","How complete and current are module materials, including dates and deadlines?","What should a course assistant refuse to do in assessed work?","How will wrong answers be reported and corrected?"],"valueHypothesis":"Interpretation: a module-grounded assistant with published limits could give students on-demand explanation and revision support with less integrity risk than open tools, if per-module accuracy is validated.","potentialEngagement":"A bounded content-readiness and evaluation engagement: audit a sample of modules, define accuracy and refusal tests, and set usage and satisfaction baselines before wider rollout.","unsupportedClaims":["Learning gains or grade improvement","Retention or continuation effects","Answer accuracy rates","Lecturer time savings"]},"engineering":{"fit":"Fits institutions on an LMS that offers a native, retrieval-grounded assistant and has reasonably structured module content.","architectureIntegration":"Assistant embedded in SurreyLearn pages, answering from module resources only, with links back to course materials and no internet search.","prerequisites":["Current, well-structured module materials","Accurate module dates and deadlines in the LMS","Data processing terms for the platform AI feature","Per-module enable or disable controls"],"constraints":["UK data-protection context; U.S. buyers must map to FERPA and state law","Vendor not named by Surrey (likely D2L, by inference)","No published accuracy or bias evaluation","Answers limited by gaps in module content"],"security":"Confirm prompt and response retention, provider access, and whether student identifiers are sent with queries; restrict retrieval so quiz and exam items are excluded.","proofOfValue":["Per-module accuracy test using real student questions","Refusal test on assessed-work requests","Check that exam and quiz content is never surfaced","Citation accuracy back to source materials"]},"delivery":{"work":"Enable the assistant module by module after content checks, publish usage guidance, and run a term-long evaluation.","dependencies":["LMS platform feature availability","Module team participation","Academic integrity policy alignment","Student communications"],"ownership":"Digital learning or LMS service owner, with module leaders responsible for content accuracy.","skillsAdoption":"Short guidance for students on verifying answers and for staff on curating module materials for retrieval.","governanceCheckpoints":["Integrity policy sign-off on permitted use","Data protection review of the platform feature","Accuracy sample review before each term","End-of-term evaluation before expansion"],"acceptanceCriteria":["Proposed: at least 90% accurate answers on a per-module sample of real questions","Proposed: 100% refusal on assessed-work completion requests in testing","Proposed: wrong-answer reports triaged within five working days","Proposed: usage and satisfaction measured against a pre-launch baseline"],"risks":["Confident wrong answers from incomplete module content","Student over-reliance in place of reading","Unclear ownership of corrections","Vendor data-processing terms not reviewed"]}},"retrievedAt":"2026-10-20T03:15:03Z","enrichedAt":"2026-10-19T03:40:00Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Could offer on-demand explanation outside teaching hours. Accessibility conformance and impact on lecturer workload are not addressed.","procurementImplications":"Interpretation: an LMS-platform AI feature rather than a separate tool, so procurement questions shift to platform contract terms, data processing and the ability to switch the feature off per module.","operatingModelImplications":"Students are told to check course materials or speak to the teaching team when answers seem wrong, which places error correction on module teams without a stated feedback loop.","sourceVerification":{"openedUrl":"https://my.surrey.ac.uk/news/meet-lumi-your-ai-powered-study-assistant","referenceExcerpt":"Unlike general AI tools, Lumi bases its answers only on the resources provided within your modules rather than searching the wider internet.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}