{"resourceId":"suazo-cognitive-offloading-autonomy-sr-20260903","versions":[{"version":"external-38bf826a85e713eb5754b1f4bfd9c3c7f87e3c3e890fca3282a538f0604e64bb","resource":{"id":"suazo-cognitive-offloading-autonomy-sr-20260903","title":"Frontiers systematic review: GenAI in HE sits between cognitive offloading and critical autonomy","organization":"Universidad Autónoma de Chile / collaborators (Suazo Galdames, Santiesteban Velázquez, Saracostti, Chaple Gil)","sector":"Higher education research (systematic review)","geography":"International higher-education evidence base (Scopus, Web of Science, EBSCOhost)","publishedAt":"Frontiers in Education, published 03 September 2026 (received 12 May 2026; accepted 21 July 2026)","publicationDate":"2026-09-03","eventDate":null,"sourceName":"Frontiers in Education","sourceLabel":"Peer-reviewed systematic review","sourceUrl":"https://www.frontiersin.org/journals/education/articles/10.3389/feduc.2026.1878665/pdf","evidenceClass":"academic-research","outcomeClass":"cautionary","topics":["knowledge-work","accessibility-workforce","governance-procurement","operating-model"],"finding":"PRISMA 2020 systematic review (PROSPERO CRD420261390452) screened 7,215 records and included 26 empirical higher-education studies (2022–May 2026 search window) on GenAI’s epistemic implications. GenAI can scaffold feedback literacy, metacognition, clarification, and self-regulated learning, but uncritical reliance was frequently associated with cognitive offloading, automation bias, superficial processing, and weakened evaluative judgment. Authors conclude consequences are mediated primarily by pedagogical design, AI literacy, and assessment practices—not the technology alone—and call for frameworks that preserve epistemic agency and cognitive autonomy.","sledRelevance":"Why-now for the 2026-09-26 edition: newly relevant scrutiny against same-edition StudentBench human-equivalence claims. Short-session GRE gains do not settle whether campus GenAI tutoring/advising preserves critical autonomy; this 26-study synthesis supplies the contradictory operating constraint student-success leaders must design for (literacy, assessment, guided use).","evidence":"Systematic review with thematic/framework synthesis; dual screening (κ=0.85); JBI-informed quality appraisal noted; HE-only empirical designs (qualitative n=11, mixed n=7, cross-sectional n=4, quasi-experimental n=2, RCT n=2).","architectureImplications":"Student-facing GenAI tools should default to verification, reflection, and source-checking steps rather than answer-final UX. Assessment systems need process evidence and oral/supervised checkpoints where evaluative judgment is the learning outcome.","governanceImplications":"Policies should move beyond bans/detection toward required AI literacy, assessment redesign, and explicit expectations for student justification of AI-supported work. Procurement should ask vendors how products discourage uncritical offloading.","securityPrivacyImplications":"Review focuses on epistemic risk more than FERPA; still implies caution against opaque AI authorities in advising/tutoring logs. Pair with campus privacy review when monitoring reliance behaviors.","caveats":"Published 2026-09-03 (before last edition) but selected for new relevance; heterogeneous designs without pooled effect sizes; English/indexed-journal bias; search closed May 2026; association language for harms, not uniform causal proof across all 26 studies.","streamIds":["student-success"],"roles":{"sales":"When buyers hear that AI tutors match human GRE gains, use this Frontiers review to reframe the deal toward durable learning design: uncritical GenAI use is repeatedly linked to cognitive offloading and weaker judgment across 26 HE studies. Sell literacy programs, guided tutor modes, and assessment redesign—not unrestricted chat seats. Bound claims to the synthesis; do not invent prevalence rates or guarantee autonomy preservation from any single product.","engineering":"Engineering should implement guided modes that require attempts, explanations, or source checks before full solutions, plus instructor-visible process traces where policy allows. Avoid one-click final-answer UX for high-stakes tasks that are meant to build judgment. Constraints: the review does not prescribe a single telemetry schema or effect size. Proof-of-value is improved unaided evaluative tasks and reduced answer-first session rates after literacy interventions.","delivery":"Delivery should sequence faculty development on AI-integrated assessment, student critical-AI workshops, and pilots that measure judgment—not only completion speed. Ownership sits with teaching-and-learning centers partnering with academic integrity offices. Risks include punitive surveillance framed as anti-offloading and ignoring the review’s finding that pedagogy and literacy mediate outcomes as much as the model choice."},"roleDetails":{"sales":{"customerProblem":"Leaders fear GenAI erodes critical thinking but lack a synthesis to justify literacy and assessment investments beside tutoring bots.","stakeholders":["teaching & learning","academic integrity","faculty senate","student success","library"],"discoveryQuestions":["Where is unguided GenAI currently the default for students?","Which assessments still only score final products?","What AI literacy is required before tool access?"],"valueHypothesis":"Pairing AI tutoring pilots with literacy and assessment redesign reduces offloading risk more credibly than seats alone.","potentialEngagement":"Campus workshop series plus redesign of 2–3 AI-exposed assessments with unaided judgment checks.","unsupportedClaims":["Pooled effect size of offloading on GPA","That detection tools restore autonomy","That all GenAI use harms learning"]},"engineering":{"fit":"Any student GenAI tutor/writing/advising assistant where answer-final UX is currently default.","architectureIntegration":"Attempt-before-hint gates, citation prompts, reflection checkpoints, optional process logs for instructors.","prerequisites":["course AI-use policy","assessment redesign partners","accessibility review of extra steps"],"constraints":["heterogeneous evidence, no single metric","search ended May 2026","equity risks if friction is poorly designed"],"security":"If logging process evidence, apply education-record controls and purpose limitation.","proofOfValue":["Rise in guided vs answer-final sessions","Unaided critical-analysis scores","Student literacy assessment completion"]},"delivery":{"work":"Launch literacy curriculum, redesign assessments for visible judgment, train faculty on guided AI pedagogy.","dependencies":["faculty time","integrity policy update","student communications"],"ownership":"Teaching-and-learning center.","skillsAdoption":"Faculty AI pedagogy; student workshops; advisor talking points on over-reliance.","governanceCheckpoints":["senate/academic board review of AI assessment norms","accessibility check on new verification steps","mid-semester student feedback on cognitive load"],"acceptanceCriteria":["Published AI literacy requirement","At least one redesigned assessment with unaided judgment component","Documented human-help alternative"],"risks":["Friction that harms equity","Moral panic overstating harms","Ignoring productive scaffolded uses the review also notes"]}},"retrievedAt":"2026-09-27T03:23:30Z","enrichedAt":"2026-09-27T03:26:58Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Critical-AI literacy and alternative assessment formats must be accessible (disability accommodations, multilingual support). Workforce: librarians, writing centers, and faculty developers become core to student-success GenAI programs—not only IT chatbot operators.","procurementImplications":"Score vendors on support for guided pedagogy and assessment integration; discount pure engagement metrics. Require documentation of how products encourage verification over offloading.","operatingModelImplications":"Treat GenAI student support as a teaching-and-learning operating model with literacy KPIs and judgment assessments, not solely a helpdesk automation program.","sourceVerification":{"openedUrl":"https://www.frontiersin.org/journals/education/articles/10.3389/feduc.2026.1878665/pdf","referenceExcerpt":"A total of 26 studies were included in the final synthesis.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}