{"resourceId":"cepr-genai-learning-penalty-chinese-secondary-2026","versions":[{"version":"external-48f280f625e7ba2d59184e70d126d9fae9d5cdb8574ee9c3c02338c39a20b9ca","resource":{"id":"cepr-genai-learning-penalty-chinese-secondary-2026","title":"CEPR/NBER evidence: GenAI raises homework scores but cuts closed-book exam learning, concentrated in homework outsourcing","organization":"CEPR Discussion Paper / NBER conference paper (Strömberg, Lei, Wu)","sector":"K-12 academic economics research (secondary education)","geography":"China (grades 7–12); transferability limits for U.S. districts","publishedAt":"June 2, 2026 (CEPR DP21577)","publicationDate":"2026-06-02","eventDate":null,"sourceName":"The Generative AI Learning Penalty: Evidence from Chinese Secondary Education","sourceLabel":"Academic research (difference-in-differences panel study)","sourceUrl":"https://conference.nber.org/conf_papers/f240545.pdf","evidenceClass":"academic-research","outcomeClass":"cautionary","topics":["knowledge-work","operating-model","accessibility-workforce"],"finding":"Using 30 months of panel data on 26,811 Chinese students in grades 7–12, staggered GenAI adoption raises homework scores ~18% and cuts completion time ~30%, but lowers monthly closed-book exam scores ~20% within six months; high-stakes entrance exams fall ~18–24% with the full penalty emerging after about two years. Losses concentrate among roughly 80% of AI users whose behavior matches homework outsourcing (very short completion time with high homework scores); users who keep similar time-on-task see small losses. Why now: U.S. districts debating bans on AI-completed assignments (e.g., APS draft; Montgomery PTA concerns about removing mental struggle) need measured evidence that assisted homework productivity is not retained learning.","sledRelevance":"Gives SLED buyers a quantitative caution against treating homework speed/score gains as proof of learning, and supports instructional designs that preserve effortful practice and off-tool assessment.","evidence":"Opened the NBER conference PDF (full paper) and the CEPR DP21577 abstract page. Verified abstract statistics (n=26,811; 18%/30%/20%; entrance-exam 18 and 24%; ~80% outsourcing concentration) and ethics approval note. CEPR hosted PDF URL returned an HTML paywall/login page, so the conference PDF is the verification object.","architectureImplications":"If product analytics celebrate homework score/time gains, add off-tool or closed-book checks and time-on-task monitors; discourage answer-complete modes for graded homework.","governanceImplications":"Academic integrity and AI-use policies should distinguish tutoring-with-effort from outsourcing; evaluate pilots on unaided assessments, not homework throughput alone.","securityPrivacyImplications":"Indirect: outsourcing patterns may push students to consumer tools outside district DPAs—reinforce approved-tool boundaries when homework AI is allowed.","caveats":"China secondary context; DiD/staggered adoption not a U.S. RCT; college entrance exams are out of pure K-12 scope except as long-run learning signal; transferability to U.S. curricula and tool ecosystems is limited and must be stated.","streamIds":["k12"],"roles":{"sales":"Sales framing: stop leading with homework completion speed. This evidence shows assisted homework gains can coincide with closed-book learning losses when students outsource work. Sell scaffolds that keep time-on-task and require unaided checks—and be explicit that Chinese secondary DiD results are directional caution for U.S. K-12, not a guaranteed local effect size.","engineering":"Engineering framing: instrument for outsourcing risk—pair homework assistance with minimum productive struggle, session time floors, and post-AI unaided items. Disable answer-reveal defaults on graded homework. Report learning with off-tool measures alongside homework analytics so dashboards cannot green-wash outsourcing.","delivery":"Delivery framing: redesign pilots so primary endpoints are closed-book or delayed unaided assessments, with homework metrics secondary. Train teachers to spot ultra-fast high-scoring AI homework patterns and to require process evidence. Acceptance is a pre-registered unaided outcome, not faster packet completion."},"roleDetails":{"sales":{"customerProblem":"Districts may misread AI homework score/time gains as learning success while unaided performance declines under outsourcing.","stakeholders":["Curriculum chiefs","Principals","Teachers","Assessment leads","Boards"],"discoveryQuestions":["What unaided assessments will judge your AI homework pilot?","Can the tool prevent full-answer outsourcing on graded work?","How will you detect ultra-fast high-score homework patterns?","Are consumer chatbots blocked for assignment completion?"],"valueHypothesis":"Effort-preserving AI homework design with unaided endpoints reduces the risk of productivity-without-learning that this study documents.","potentialEngagement":"5-week homework-AI redesign: analytics review, outsourcing heuristics, and an unaided assessment pilot protocol.","unsupportedClaims":["That U.S. districts will see the same 20% exam penalty","Guaranteed learning gains from any AI tutor mode","That banning AI alone fixes outsourcing on consumer devices"]},"engineering":{"fit":"Fit for homework/practice products that can enforce scaffolding and unaided checks; poor fit for answer-complete chatbots on graded work.","architectureIntegration":"Homework AI behind LMS with process logs; delayed unaided quizzes; time-on-task signals; optional AI-off assignments.","prerequisites":["Baseline unaided assessments","Teacher workflow for process review","Filtering of consumer GenAI where policy requires"],"constraints":["Evidence from China grades 7–12","Observational DiD, not U.S. RCT","Students may bypass on personal devices"],"security":"Keep assignment assistance on approved enterprise tools under DPA; log enough process data for integrity review without over-collecting.","proofOfValue":["Pre/post unaided assessment in pilot sample","Share of sessions flagged as outsourcing-like","Homework time distribution vs non-AI peers"]},"delivery":{"work":"Set unaided endpoints, configure anti-outsourcing modes, train teachers on process evidence, and report dual metrics (assisted vs unaided).","dependencies":["Assessment team","LMS admin","Academic integrity policy"],"ownership":"Chief academic officer owns learning endpoints; principals own classroom enforcement.","skillsAdoption":"Teachers need PD to distinguish tutoring from outsourcing and to design AI-off checks.","governanceCheckpoints":["Pilot pre-registration of unaided outcomes","Midpoint outsourcing-pattern review","Go/no-go on expanding homework AI"],"acceptanceCriteria":["Primary outcome is unaided/closed-book measure","Documented response when outsourcing patterns spike","Consumer GenAI assignment path addressed in AUP"],"risks":["Over-generalizing China effect sizes","Teachers equated homework scores with mastery","Equity issues if only some students have stealth AI access"]}},"retrievedAt":"2026-09-19T03:09:40Z","enrichedAt":"2026-09-19T03:13:15Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Protecting productive struggle may conflict with legitimate assistive uses for some disabilities—pair integrity rules with IEP-aligned accommodations rather than blanket bans.","procurementImplications":"Score vendors on unaided-learning evaluation plans and anti-outsourcing defaults, not homework throughput dashboards alone.","operatingModelImplications":"Shift AI homework pilots into the same operating cadence as assessment: pre-register endpoints, monitor process metrics, and stop when unaided learning falls.","sourceVerification":{"openedUrl":"https://conference.nber.org/conf_papers/f240545.pdf","referenceExcerpt":"The learning losses are concentrated among roughly 80% of AI users whose behavior is consistent with homework outsourcing","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}