{"resourceId":"korea-molit-ai-preliminary-permit-assessment-pilot-5-localities-20261104","versions":[{"version":"external-e6e8e169328ed99b1cc2fad627cdd3de33b2d6d602637a8c01ee753e64eee949","resource":{"id":"korea-molit-ai-preliminary-permit-assessment-pilot-5-localities-20261104","title":"South Korea's land ministry unveils a multi-agent AI service that pre-assesses whether land-development permits are possible from laws and local ordinances; pilots in five localities from January 2027, 18 from June","organization":"Ministry of Land, Infrastructure and Transport (MOLIT) and Ministry of Science and ICT, with a BI Matrix-led consortium (WAVUS, ICT Way)","sector":"Permitting and land-development approvals","geography":"South Korea (pilots in Gyeyang-gu Incheon, Eumseong, Asan, Gyeongsan, Yeongcheon)","publishedAt":"2026-11-04","publicationDate":"2026-11-04","eventDate":"2026-11-04","sourceName":"Financial News (fnnews, English edition)","sourceLabel":"“Can I build a house on this land?” ... AI conducts a preliminary permit assessment","sourceUrl":"https://en.fnnews.com/news/202611040817513022","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["developers-agents","governance-procurement","operating-model"],"finding":"Financial News reports that on Nov. 4, 2026 MOLIT briefed about 130 participants on an AI service that makes a preliminary assessment of whether land-development permits and approvals are possible, based on land-development laws and local government ordinances. It is one of MSIT's 'Top 10 AI Projects for People's Livelihoods'; development began in April 2026 by a BI Matrix-led consortium and runs to December 2027. An orchestrator coordinating multiple AI agents and a test environment are complete, and the service interprets everyday-language questions. Pilots start in January 2027 in five localities; the ministries will establish the reliability of assessments before expanding to 18 local governments from June 2027, then add mobile and an official-facing tool in December 2027.","sledRelevance":"A national-scale, agent-based pre-application permit screen with an explicit reliability gate before expansion, a useful comparator for U.S. pre-application tools (Denver StartSmart, Burlington AI PreCheck, Baguio zoning checker already archived). Dated Nov. 4, after CE's last permitting coverage, and first-time in the archive; included because its staged reliability plan is new evidence on how to roll out permit-screening AI.","evidence":"News report of a government briefing; describes completed components and a pilot schedule; no accuracy results yet, since reliability is to be established during pilots.","architectureImplications":"Multi-agent orchestrator over statutes, local ordinances and geospatial data (WAVUS, ICT Way), with intent recognition for plain-language questions.","governanceImplications":"Reliability to be established in five pilot areas before expansion; outputs are preliminary assessments, not permits; the legal weight of an assessment is not described.","securityPrivacyImplications":"Parcel-level queries may reveal development intentions; data handling is not described.","caveats":"Secondary news report; pre-pilot; no accuracy data; Korean land-use law and national ordinance data structures differ from fragmented U.S. zoning codes.","streamIds":["citizen-engagement"],"roles":{"sales":"Sales framing: Korea is funding a national pre-application screen that tells residents whether a permit is even possible before they apply, and it is gating expansion on reliability. Discovery should ask how many inquiries are 'can I build here?' questions, whether local code is machine-readable, and who would stand behind a preliminary answer. A bounded engagement is a code-readiness and test-set study for a pre-application screen. Do not claim accuracy or time savings; the pilots have not started.","engineering":"Engineering framing: this is a multi-agent orchestrator over statutes, local ordinances and parcel data. The hard prerequisite is structured, current ordinance data for each jurisdiction. Constraints include unclear legal weight of assessments and fragmented U.S. zoning codes. Proof of value should compare AI assessments with planner determinations on a sample of real parcels, measure disagreement by rule type, and test plain-language intent recognition on actual resident questions.","delivery":"Delivery framing: the ministries are staging rollout to five pilots, then 18 localities, with reliability checked first. A U.S. equivalent would need planners to own the rules and review disagreements. Proposed acceptance criteria: agreement with planner determinations above an agreed rate on a parcel sample, a disclaimer that assessments are not approvals, and an expansion gate tied to measured reliability. Budget for ordinance digitisation per jurisdiction."},"roleDetails":{"sales":{"customerProblem":"Residents and developers who cannot tell whether a parcel can be developed without paying for advice or applying.","stakeholders":["Planning director","Permit centre manager","GIS lead","City attorney"],"discoveryQuestions":["How many inquiries are feasibility questions?","Is the zoning code machine-readable?","Who stands behind a preliminary answer?"],"valueHypothesis":"A pre-application screen may reduce futile applications and staff feasibility calls; no pilot results yet.","potentialEngagement":"A code-readiness assessment and parcel test set for a pre-application feasibility screen.","unsupportedClaims":["Accuracy of assessments","Faster approvals","Legal reliance on AI answers"]},"engineering":{"fit":"Fits jurisdictions with digitised zoning and parcel GIS data.","architectureIntegration":"Multi-agent orchestrator over codified ordinances, statutes and parcel geospatial layers with plain-language intent recognition.","prerequisites":["Machine-readable ordinances","Parcel GIS layers","Planner-labelled test parcels"],"constraints":["Fragmented U.S. zoning codes","Unclear legal status of assessments","Korean-law specifics"],"security":"Log queries, avoid exposing applicant intent, and keep assessments clearly non-binding.","proofOfValue":["Agreement with planner determinations","Disagreement analysis by rule type","Intent recognition on real questions"]},"delivery":{"work":"Digitise ordinances, build a parcel test set, and pilot in a few areas before expanding.","dependencies":["Planning staff","GIS team","Ordinance data"],"ownership":"Planning department with GIS as data owner.","skillsAdoption":"Train planners to label test parcels and review AI disagreements.","governanceCheckpoints":["Reliability gate before expansion","Legal review of disclaimers","Ordinance update process"],"acceptanceCriteria":["Agreement rate above an agreed target on test parcels","Disclaimer shown on every assessment","Expansion only after reliability review"],"risks":["Residents relying on wrong assessments","Stale ordinance data","Scope creep before reliability proven"]}},"retrievedAt":"2026-11-10T22:23:01-06:00","enrichedAt":"2026-11-10T22:29:03-06:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Aims to let residents check feasibility without knowing the law; an official-facing tool is planned for December 2027.","procurementImplications":"Funded through a national AI programme with an industry consortium over two years; illustrates programme-funded development rather than per-city purchase.","operatingModelImplications":"Staged rollout: 5 pilots, then 18 local governments, then mobile and staff tools, with ordinances from each locality needing to be machine-readable.","sourceVerification":{"openedUrl":"https://en.fnnews.com/news/202611040817513022","referenceExcerpt":"establish the reliability of its assessments before expanding the pilot program to 18 local governments starting in June 2027.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}