{"resourceId":"new-york-psc-utilities-ai-inventory-reporting-20260917","versions":[{"version":"external-29d99bb27796e04bddde5c881779e4a4cb9358072e948718249acb2f4b48b57e","resource":{"id":"new-york-psc-utilities-ai-inventory-reporting-20260917","title":"New York PSC orders utility AI inventories and semiannual reporting (Case 26-M-0552)","organization":"New York State Public Service Commission / Department of Public Service","sector":"State utility regulation — critical infrastructure AI oversight","geography":"New York (investor-owned utilities; other jurisdictions transfer with regulatory differences)","publishedAt":"2026-09-17","publicationDate":"2026-09-17","eventDate":"2026-09-17","sourceName":"WXXI News / New York Public News Network (with Daily Energy Insider hearing detail)","sourceLabel":"Independent reporting of PSC action","sourceUrl":"https://www.wxxinews.org/new-york-public-news-network/2026-09-17/new-york-to-begin-monitoring-utilities-use-of-ai","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["governance-procurement","infrastructure","data-security","operating-model"],"finding":"On Sept. 17, 2026, the NY PSC voted to require regulated utilities to inventory AI use—including generative and machine-learning systems used for inspections, outage modeling, customer service, and analytics—and to file governing policies, with an initial inventory due in 60 days and semiannual reports thereafter (Case 26-M-0552 per hearing coverage). Chair Christian acknowledges possible net benefits while staff highlight misconfiguration, hallucination, privacy, and cybersecurity risks in critical infrastructure. No measured reliability outcomes yet.","sledRelevance":"State utility commissions and energy CIOs are a transferable oversight pattern for other regulated critical infrastructure: start with inventories and policy disclosures before prescribing technical standards. Liability for AI-driven ratepayer harm was raised but deferred.","evidence":"WXXI: PSC vote for semiannual AI reports; 60-day first filing; Christian and DiBari quotes; National Grid will provide inventory. Daily Energy Insider: unanimous vote; Case 26-M-0552; inventory covers internally built and externally procured AI; risks list; commissioners asked about standardization, liability, and agentic AI currency.","architectureImplications":"Utilities will need durable AI asset inventories spanning vendor SaaS, internal models, and embedded inspection analytics, plus change-control so novel agentic uses update filings continuously as staff suggested.","governanceImplications":"Creates a regulatory discovery proceeding rather than immediate technical standards. Expect later standardization and prudence/liability questions once inventories arrive.","securityPrivacyImplications":"Explicitly frames AI misconfiguration and fabricated outputs as a new critical-infrastructure risk category for commission evaluation; privacy and cybersecurity attacks listed among staff concerns.","caveats":"Primary signed order PDF not opened this run; facts triangulated from WXXI and Daily Energy Insider. Benefits remain acknowledged as possible, not measured. Separate affordability-index proceeding is not an AI finding.","streamIds":["state-government"],"roles":{"sales":"Sales framing: New York just forced regulated utilities to produce AI inventories and policy disclosures on a 60-day clock. Engage PSC-facing utility CIOs, OT/IT risk owners, and rate-case counsel who must catalogue inspection, outage, and customer-service AI before standards land. Position a bounded inventory-and-controls readiness engagement—not guaranteed bill savings or outage reductions. Unsupported: that reporting alone improves grid reliability.","engineering":"Engineering framing: Fit is enterprise AI discovery across OT analytics, outage models, and customer systems—including procured GenAI—with policy/procedure evidence packs for regulators. Integrate continuous inventory updates for novel agentic uses. Prerequisites: agreed AI definitions (staff cited federal and NYS ITS definitions), data-flow maps, and vendor contract visibility. Constraints: New York PSC jurisdiction; liability allocation for AI errors not decided in this order.","delivery":"Delivery framing: Work is a 60-day initial inventory plus recurring semiannual report machinery, with named owners per use case and policy attestation. Ownership typically sits with utility compliance/regulatory affairs plus CIO/CISO. Acceptance should include complete internal-plus-procured coverage, mapped risks, and versioned policy links—not a slide deck. Risks: under-reporting shadow AI; inventory staleness as agentic tools arrive; future prudence findings if inventories reveal weak controls."},"roleDetails":{"sales":{"customerProblem":"Regulated utilities lack a complete, commission-ready inventory of AI systems touching inspections, outages, and customers—and now face timed disclosure duties.","stakeholders":["Utility CIO/CISO","Regulatory affairs / PSC counsel","OT operations leaders","Vendor AI providers"],"discoveryQuestions":["Which AI systems are internally built versus procured, and who attests each entry?","What policies govern generative versus classical ML uses today?","How will agentic or novel uses trigger inventory updates between semiannual filings?","Where could AI errors create ratepayer harm under existing prudence standards?"],"valueHypothesis":"A complete, continuously maintained AI inventory with linked policies may reduce regulatory surprise and prepare utilities for later standards—without proving reliability gains yet.","potentialEngagement":"6–8 week Case 26-M-0552 readiness: AI discovery across OT/IT, policy mapping, draft initial filing, and update workflow design.","unsupportedClaims":["Guaranteed outage reduction or customer-service improvement from reporting","That inventories equal approved AI safety cases","Statewide applicability outside NY PSC jurisdiction"]},"engineering":{"fit":"Regulatory AI asset inventory and control-evidence architecture for investor-owned utilities.","architectureIntegration":"Connect CMDB/OT asset systems, vendor SaaS inventories, model registries, and policy repositories; support continuous updates for new agentic uses.","prerequisites":["Scoped AI definition matching staff guidance","Cross-OT/IT discovery authority","Vendor contract clauses revealing subprocessors and model roles"],"constraints":["60-day initial filing clock","Liability questions deferred by commission counsel","Primary order PDF not independently opened this run"],"security":"Inventory itself is sensitive; protect details that would aid attackers while meeting disclosure duties.","proofOfValue":["Initial filing covering internal and procured systems","Traceable policy links per use case","Documented update path for novel AI between semiannual reports"]},"delivery":{"work":"Build inventory, collect governing policies/procedures, file within 60 days, and operationalize semiannual refresh plus continuous novel-use updates.","dependencies":["DPS filing portal/process","Business-unit SMEs for inspections/outage/customer AI","Legal review of disclosures"],"ownership":"Utility regulatory compliance lead with CIO/CISO co-ownership of inventory accuracy.","skillsAdoption":"Need OT analytics literacy and GenAI governance skills among compliance staff—not only traditional cybersecurity checklists.","governanceCheckpoints":["60-day initial inventory filing","First semiannual refresh","Internal review when agentic use cases appear"],"acceptanceCriteria":["Inventory includes internal and externally procured AI","Each entry cites governing policy/procedure","Named attestation owner and refresh cadence"],"risks":["Incomplete discovery / shadow AI","Over-disclosure of sensitive architecture","Future prudence exposure if weak controls are inventoried without remediation plans"]}},"retrievedAt":"2026-09-18T22:12:45-05:00","enrichedAt":"2026-09-18T22:17:59.954515-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Customer-service AI uses may affect accessibility of utility interactions; the order requires disclosure of use types but does not set accessibility performance metrics.","procurementImplications":"Externally procured AI is explicitly in scope—vendor contracts should anticipate inventory cooperation, subprocessor transparency, and future commission standards.","operatingModelImplications":"Moves utility AI from informal adoption to regulated discovery with recurring reporting, likely forcing cross-functional ownership among OT, IT, customer operations, and regulatory affairs.","sourceVerification":{"openedUrl":"https://www.wxxinews.org/new-york-public-news-network/2026-09-17/new-york-to-begin-monitoring-utilities-use-of-ai","referenceExcerpt":"The New York State Public Service Commission voted Thursday to order the state’s utilities to file semiannual reports on the technology.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}