{"resourceId":"epa-advisory-ai-providers-requesting-water-scada-logs-2026","versions":[{"version":"external-9553543e4d00a50e96b75709344175a8067cf7eea82189a9b558f45c26062f6f","resource":{"id":"epa-advisory-ai-providers-requesting-water-scada-logs-2026","title":"EPA advisory: AI service providers requesting water/wastewater SCADA and operational logs via public records","organization":"U.S. Environmental Protection Agency Office of Water (Emergency Response and Cybersecurity)","sector":"Water and wastewater utilities / federal advisory","geography":"United States","publishedAt":"2026-06-05","publicationDate":"2026-06-05","eventDate":null,"sourceName":"EPA Office of Water GovDelivery bulletin","sourceLabel":"EPA Advisory: Protecting Sensitive Operational Information in Water and Wastewater Systems","sourceUrl":"https://content.govdelivery.com/accounts/USEPAOGWDW/bulletins/41a983a","evidenceClass":"standards-guidance","outcomeClass":"cautionary","topics":["data-security","governance-procurement","developers-agents"],"finding":"EPA’s 5 June 2026 advisory warns water and wastewater systems about disclosing sensitive operational information after multiple public-records requests from AI service providers, as reported by WaterISAC and AWWA. In at least one instance, an AI provider requested all SCADA logs for 2026 for treatment, distribution and wastewater systems, plus daily/hourly historical logs for flows, water-quality baselines, reservoir/tank levels and energy or equipment run-times. EPA concurs that aggregating such operational data across utilities could reveal patterns and vulnerabilities usable to disrupt service. Utilities are urged to make risk-informed decisions consistent with state information-protection laws. Why now: NARUC’s 8-11 Nov WC-1 draft seeks FOIA protections for cyber-incident sharing and WRRO minimums amid the same OT data-exposure risk surface.","sledRelevance":"Direct OT/AI security control for U.S. SLED water utilities: treat AI vendor data requests for SCADA/process historian extracts as high-risk even when framed as research or model training. Newly relevant alongside this week’s NARUC WRRO/CISA draft.","evidence":"Opened the EPA GovDelivery bulletin text dated 06/05/2026 09:50 AM EDT from Office of Water Emergency Response and Cybersecurity.","architectureImplications":"Segment and classify SCADA/historian exports; require legal and CISO review before fulfilling broad AI-related records requests; prefer redacted aggregates over raw logs.","governanceImplications":"Align public-records response playbooks with cyber risk reviews; coordinate with counsel on state disclosure laws versus security exceptions.","securityPrivacyImplications":"Bulk SCADA and process logs enable reconnaissance; aggregation across utilities amplifies sector risk as EPA states.","caveats":"Advisory dated June 2026 (older than last run) included for why-now linkage to NARUC WC-1 vote week. Does not name the requesting provider or quantify how many utilities received requests.","streamIds":["utility-management"],"roles":{"sales":"Sales framing: use the EPA advisory to open security-minded discovery when utilities evaluate AI vendors that want operational data for training or analytics. Position engagements around data-minimisation, redaction and lawful public-records handling—not around granting bulk SCADA access. Tie the conversation to this week’s NARUC WC-1 FOIA and WRRO themes without claiming the advisory creates new federal mandates.","engineering":"Engineering framing: classify SCADA, historian and process KPIs as sensitive OT data; design AI pilots that use synthetic or tightly scoped features rather than year-long raw logs. Prerequisites include an information-classification scheme and a joint legal/CISO review gate. Proof of value is a documented request-triage workflow that blocks unconstrained log dumps, and record those bounds in the engagement brief.","delivery":"Delivery framing: implement a request-triage SOP with counsel and CISO sign-off, a redaction checklist, and logging of AI-related records requests for after-action review. Ownership sits with the utility records officer and CISO jointly. Risks include over-disclosure under state FOIA pressure and under-sharing that still needs a clear written rationale on the file, and record those bounds in the engagement brief."},"roleDetails":{"sales":{"customerProblem":"AI vendors and public-records requesters seeking bulk OT logs create cyber exposure.","stakeholders":["CISO","Utility counsel/records officer","Plant superintendent","State primacy agency"],"discoveryQuestions":["Have you received AI-related SCADA or historian requests?","Who approves public-records releases involving OT data?","What redaction standards exist today?"],"valueHypothesis":"A triage and minimisation playbook reduces breach-via-FOIA risk without blocking legitimate analytics.","potentialEngagement":"OT data-classification and request-triage workshop aligned to EPA advisory.","unsupportedClaims":["Federal ban on all AI data requests","Named threat actor attribution","Automatic FOIA exemption without state law"]},"engineering":{"fit":"Controls and data-governance overlay—not an AI model deploy.","architectureIntegration":"Classify historian/SCADA exports; provide feature stores or aggregates instead of raw logs for AI POCs.","prerequisites":["Asset/data inventory","Legal review channel","Logging of records requests"],"constraints":["State public-records variation","June advisory date","No technical control catalogue in the bulletin"],"security":"Deny bulk multi-system SCADA log releases by default; require risk review.","proofOfValue":["Triaged request log","Zero unconstrained SCADA dumps in pilot window","Documented redaction examples"]},"delivery":{"work":"Draft SOP, train records/cyber staff, tabletop an AI vendor request scenario.","dependencies":["Counsel availability","WaterISAC/AWWA guidance uptake","Executive sponsorship"],"ownership":"Records officer executes releases; CISO owns cyber risk decision.","skillsAdoption":"Short training for customer service/records staff who first see requests.","governanceCheckpoints":["SOP legal review","First live request after-action","Annual playbook refresh"],"acceptanceCriteria":["Signed SOP","Request log template in use","Escalation path tested"],"risks":["Political pressure to over-release","Inconsistent state law interpretation","Shadow AI POCs bypassing SOP"]}},"retrievedAt":"2026-11-06T22:05:03.109776-06:00","enrichedAt":"2026-11-06T22:14:20-06:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Small utilities without counsel or CISO capacity need shared playbooks via WaterISAC/state associations.","procurementImplications":"AI vendor contracts and POCs must forbid unconstrained SCADA-log collection; require data-minimisation clauses.","operatingModelImplications":"Records and cyber teams must jointly triage AI-related information requests before release.","peerReview":"unknown","sourceVerification":{"openedUrl":"https://content.govdelivery.com/accounts/USEPAOGWDW/bulletins/41a983a","referenceExcerpt":"an AI service provider requested “all SCADA logs for 2026 for all treatment, distribution and wastewater management systems.”","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}