{"resourceId":"longbeach-commscoach-privacy-report-card-2026","versions":[{"version":"external-a54993ec38c17659946878691f2d3fe42ab3624f6ecd5c17e49d0cd6414bf97e","resource":{"id":"longbeach-commscoach-privacy-report-card-2026","title":"Long Beach publishes CommsCoach privacy report card: 100% QA evals, GovCloud, 24-hour unmapped deletion","organization":"City of Long Beach Disaster Preparedness and Emergency Communications (DPEC)","sector":"911 / ECC quality assurance governance","geography":"Long Beach, California, United States","publishedAt":"2026","publicationDate":null,"eventDate":null,"sourceName":"City of Long Beach Smart City / privacy evaluation package","sourceLabel":"Municipal privacy governance publication","sourceUrl":"https://www.longbeach.gov/globalassets/smart-city/media-library/documents/colb-commscoach-privacy-report-card-package","evidenceClass":"standards-guidance","outcomeClass":"cautionary","topics":["data-security","governance-procurement","operating-model","knowledge-work"],"finding":"Long Beach’s official CommsCoach privacy report card and PIA package document fully automated QA of ECC calls/radio/CAD, claim automatic evaluation of 100% of 911 calls, radio, and CAD records, process in AWS GovCloud with CJIS/SOC2 framing, delete unmapped GovCloud data within 24 hours, retain mapped workstation data for simulator/QA duration, and acknowledge OpenAI model bias risk while using binary yes/no evaluation criteria. Distinct from the April Hoodline feature already archived: this is the city’s governance artifact with retention and prohibited-use detail.","sledRelevance":"Gives SLED ECCs a concrete U.S. municipal template for post-call AI QA privacy controls—especially retention clocks and bans on unauthorized employee monitoring—when comparing vendors after CommsCoach entered the archive on 2026-09-18.","evidence":"Privacy report card retention section; PIA statements on 100% automated evaluations; GovCloud encryption; OpenAI bias link; prohibited uses including unauthorized employee monitoring.","architectureImplications":"Logging-recorder + CAD export into GovCloud serverless QA; mapped vs unmapped data paths with different retention; OpenAI models behind binary rubric scoring.","governanceImplications":"Public report card + PIA/PUP process; notify privacy office; define coaching vs discipline boundaries; document generative-AI guidance adherence.","securityPrivacyImplications":"24-hour deletion of unmapped audio/data; GovCloud encryption at rest/in transit; CCPA/Delete Act references; residual model-bias risk disclosed via OpenAI article link.","caveats":"Package date not clearly stamped as a single ISO day on the report card face; performance benefit claims are purpose statements, not outcome trials; OpenAI dependency is a supply-chain risk.","streamIds":["emergency-services"],"roles":{"sales":"Lead with governance readiness: agencies that want ECC QA AI should expect a public privacy report card, GovCloud path, and a 24-hour unmapped deletion clock—not just a demo score. Use Long Beach to show what “responsible QA AI” paperwork looks like. Do not claim 100% evaluation equals better clinical or answer-time outcomes; that is not evidenced here.","engineering":"Replicate mapped/unmapped data handling: anything not tied to a workstation identity deletes in 24 hours; mapped training/QA data follows longer but bounded retention. Prefer GovCloud/CJIS-capable regions. Keep OpenAI (or equivalent) calls behind a rubric that is auditable yes/no, and log overrides. Proof: deletion job evidence, access audits, and sample human re-score agreement.","delivery":"Stand up PIA/PUP, configure retention jobs, train supervisors on coaching-only use, publish the report card. Ownership: DPEC-equivalent + city privacy office. Acceptance: evidenced 24-hour unmapped deletion, 100% queue coverage config, written ban on covert disciplinary monitoring. Risks: model bias, retention misconfiguration, scope creep into live triage."},"roleDetails":{"sales":{"customerProblem":"ECCs want AI QA coverage but lack a public privacy/retention pattern that councils and unions will accept.","stakeholders":["ECC directors","City privacy/IT","Unions","City attorney","Council"],"discoveryQuestions":["Is there a public PIA?","Unmapped deletion SLA?","GovCloud/CJIS?","Coaching vs discipline policy?"],"valueHypothesis":"If a city can show GovCloud + 24h unmapped deletion + coaching-only policy, QA AI procurement clears privacy gates faster.","potentialEngagement":"Privacy report-card and retention-control design workshop for ECC QA AI.","unsupportedClaims":["Clinical outcome improvement","Bias-free OpenAI scoring","Safe live 911 AI","FTE elimination guarantees"]},"engineering":{"fit":"Post-call automated QA on recorder+CAD data in GovCloud with strict retention bifurcation.","architectureIntegration":"Ingest recorder/CAD; map to workstation identities; score via binary rubric on LLM backend; delete unmapped blobs in 24h.","prerequisites":["GovCloud account","Recorder/CAD feeds","Identity mapping","PIA/PUP"],"constraints":["OpenAI bias/supply chain","Binary rubric may miss nuance","Not for live answering"],"security":"Encryption at rest/in transit; access audit; CJIS/SOC2 controls; retention jobs monitored.","proofOfValue":["Deletion job logs","Access audit samples","Human re-score agreement","Override rate"]},"delivery":{"work":"Complete PIA, configure retention, train coaching use, publish report card, monitor deletion jobs.","dependencies":["Vendor GovWorx/CommsCoach","AWS GovCloud","City privacy office","Union notice if required"],"ownership":"DPEC with Technology & Innovation privacy office oversight.","skillsAdoption":"Supervisors use AI scores for coaching conversations, not silent discipline.","governanceCheckpoints":["PIA approval","Quarterly retention audit","Bias complaint pathway"],"acceptanceCriteria":["24h unmapped deletion evidenced","Public report card posted","Written prohibited-use list enforced"],"risks":["Retention job failure","Scope creep to live AI","Union grievances over monitoring"]}},"retrievedAt":"2026-09-19T22:11:22.863731-05:00","enrichedAt":"2026-09-19T22:15:56-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Binary scoring and simulator features can broaden QA coverage but may miss nuanced communication quality affecting LEP callers unless rubrics include accessibility items.","procurementImplications":"Require GovCloud/CJIS, deletion SLAs, and prohibited-use clauses in ECC AI QA contracts.","operatingModelImplications":"Shifts QA from sparse manual sampling toward near-census review with human coaching interpretation.","sourceVerification":{"openedUrl":"https://www.longbeach.gov/globalassets/smart-city/media-library/documents/colb-commscoach-privacy-report-card-package","referenceExcerpt":"Any data that is sent to AWS GovCloud that is not mapped to a 911 workstation is deleted within 24 hours.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}