{"resourceId":"ct-ag-occ-pura-ai-utility-probe-petition-2026","versions":[{"version":"external-47a87326164c30d9baac85a70e8293e59bf72d7a69c6c88fd52a86418f5ac189","resource":{"id":"ct-ag-occ-pura-ai-utility-probe-petition-2026","title":"Attorney General Tong, Consumer Counsel Coleman Petition PURA for Probe of Artificial Intelligence Use by Public Utilities","organization":"Connecticut Office of the Attorney General and Office of Consumer Counsel","sector":"State utility regulation (electric, gas, water)","geography":"Connecticut, United States","publishedAt":"13 Oct 2026 (press release)","publicationDate":"2026-10-13","eventDate":"2026-10-13","sourceName":"Connecticut Office of the Attorney General","sourceLabel":"State government press release","sourceUrl":"https://portal.ct.gov/ag/press-releases/2026-press-releases/pura-petition-for-probe-of-artificial-intelligence-use-by-public-utilities","evidenceClass":"government-evaluation","outcomeClass":"cautionary","topics":["governance-procurement","data-security","operating-model"],"finding":"Connecticut's Attorney General and Consumer Counsel petitioned PURA to open a comprehensive review of utility AI use, asking that utilities report all AI systems, their objectives, level of human oversight and governing policies, and that PURA review data-security and cybersecurity protections and breaches. Requested scope covers storm response, customer systems, asset management, load and water-demand forecasting, procurement, grid/plant/water treatment operations, third-party vendors and AMI. The petitioners cite risks including hallucination and bias, and reports that conventional weather models outperformed AI models for a February nor'easter.","sledRelevance":"Signals that state regulators and consumer advocates may require AI inventories and human-oversight disclosures from electric and water utilities, following New York's September 2026 order already in the SLED archive.","evidence":"Advocacy petition and press release; it states concerns and requested scope but presents no new measured findings. Whether PURA opens the docket was not established at retrieval.","architectureImplications":"Utilities would need an inventory of AI systems across OT, AMI, customer systems and vendor platforms, including embedded vendor AI.","governanceImplications":"Expect regulator requests for AI use-case inventories, oversight levels, and policies benchmarked against industry standards; third-party vendor AI is explicitly in scope.","securityPrivacyImplications":"Requests review of AI and non-AI cybersecurity protections and breaches, and protection of sensitive customer information, including AMI data.","caveats":"Petition only; PURA action and timing unknown. Claims about AI forecast performance are cited by the petitioners, not independently evidenced in the release. Named utilities are investor-owned (Avangrid, Eversource), though water utilities are in requested scope. Classed government-evaluation as the closest schema class; it contains no evaluative findings.","streamIds":["utility-management","state-government"],"roles":{"sales":"Sales framing: this petition is a governance trigger, not a buying signal for AI tools. Connecticut advocates want utilities to disclose every AI system, its purpose and its human oversight, including vendor-embedded AI and AMI analytics. For municipal and regulated utility leaders in the Northeast, open with a readiness conversation about whether they could produce such an inventory today. Position a bounded AI inventory and policy gap assessment; do not imply the review has been opened or that compliance requirements are final.","engineering":"Engineering framing: the requested scope spans OT operations, water treatment, AMI, customer systems and third-party vendors, so discovery must reach beyond IT-owned chatbots. Engineers should map where models run, what data they consume, how outputs reach operators, and where human approval exists. A proof of value is a sampled inventory across two or three domains, tagging each system by oversight level and data sensitivity. Jurisdictional limits apply because PURA has not yet acted on the petition.","delivery":"Delivery framing: ownership sits with the utility's regulatory affairs lead in partnership with the CIO and CISO. Work includes building a living AI register, documenting oversight levels, and aligning policies to a recognized framework such as NIST AI RMF. Governance checkpoints should include legal review of filings and vendor attestations. Risks include incomplete discovery of embedded vendor AI, inconsistent definitions, and filing disclosures that create security exposure if OT details are over-shared publicly."},"roleDetails":{"sales":{"customerProblem":"Utilities may soon need to disclose all AI use, oversight and safeguards to regulators but lack a complete inventory.","stakeholders":["Regulatory affairs director","CIO","CISO","General counsel","Operations leadership"],"discoveryQuestions":["Could you list every AI system in operations, customer service and vendor platforms today?","How do you document human oversight for AI-informed decisions?","Which vendor contracts disclose embedded AI features?"],"valueHypothesis":"A prepared inventory and policy baseline may reduce scramble and disclosure risk if regulators open an inquiry; it does not guarantee regulatory approval.","potentialEngagement":"A time-boxed AI inventory and governance gap assessment against NIST AI RMF categories.","unsupportedClaims":["That PURA has opened a docket","Specific future compliance requirements","That AI forecasts are generally worse than conventional models"]},"engineering":{"fit":"Fits utilities with mixed IT/OT estates and vendor platforms carrying embedded AI.","architectureIntegration":"Inventory spans SCADA/OT analytics, AMI head-end and MDM analytics, CIS/billing, outage management and enterprise copilots.","prerequisites":["Application and asset inventories","Vendor contract access","OT network documentation"],"constraints":["Connecticut-specific; outcome of petition unknown","Sensitive OT details may need confidential filing treatment"],"security":"Inventory must classify data sensitivity and avoid publishing exploitable OT architecture details.","proofOfValue":["Sampled inventory of two domains with oversight tags","Gap map against one governance framework"]},"delivery":{"work":"Build an AI system register, oversight classification, and policy alignment package.","dependencies":["Executive sponsor","Vendor cooperation","Legal review"],"ownership":"Regulatory affairs lead with CIO/CISO.","skillsAdoption":"Staff need shared AI definitions and a lightweight intake process for new tools.","governanceCheckpoints":["Legal review before filing","CISO review of disclosure content","Annual register refresh"],"acceptanceCriteria":["Proposed: register covers all named scope areas","Proposed: each entry has owner, purpose, oversight level and data class"],"risks":["Missed embedded vendor AI","Over-disclosure of OT details","Definitions drift across departments"]}},"retrievedAt":"2026-10-29T22:06:05-05:00","enrichedAt":"2026-10-29T22:18:16-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Requires staff time across departments; human-oversight documentation may surface training needs.","procurementImplications":"Third-party vendors are named in scope; contracts may need AI disclosure and audit clauses.","operatingModelImplications":"Moves AI governance from ad hoc IT policy toward a regulator-facing register maintained by regulatory affairs.","sourceVerification":{"openedUrl":"https://portal.ct.gov/ag/press-releases/2026-press-releases/pura-petition-for-probe-of-artificial-intelligence-use-by-public-utilities","referenceExcerpt":"the extent and scope of AI use in Connecticut utility operations is unclear.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}