From the Emergency Services edition of October 2, 2026
Seattle Times: Fire Department used Corti AI on 911 medical calls without caller notice or public metrics
The Seattle Times · Independent journalism / EMS 911 triage · United States — Seattle, Washington
- Publisher
- The Seattle Times
- Original publication
- June 14, 2026
- Source retrieved
- 2026-10-05
What happened
Seattle Times reporting finds Seattle Fire used Corti AI for years on 911 medical audio—initially quality assurance, then live prompts from December 2023—to help divert some callers to a nurse line. Callers were not told AI was listening; the department lacked public outcome metrics and detailed contract accounting (about $260,000/year cited). Officials say dispatchers retain final authority.
Why it matters
Primary-source governance warning for emergency-services AI on live 911 medical audio: disclosure, surveillance review, and published safety KPIs must accompany any triage/diversion prompting—especially as other agencies scale non-emergency AI assistants and wildfire-detection AI.
Evidence and measured results
Opened Seattle Times watchdog article (June 14, 2026). Documents multi-year Corti use, live prompts since Dec 2023, absence of caller notice and public metrics, ~$260k annual cost, human-final-authority statements, and expert concerns about governance matching stakes.
Limitations and uncertainty
Older than the 2026-10-01 last published run; included because primary Times URL was not in emergency-services archive (EMS1 republication was) and remains newly relevant as wildfire/PSAP AI adoption expands. Clinical benefit/harm still unmeasured in the article.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-04; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Agencies want 911 triage AI without repeating Seattle’s disclosure and metrics failures.
- Stakeholders
- Fire/EMS chiefs
- City privacy/surveillance boards
- Dispatchers
- Public auditors
- Discovery questions
- Will callers be notified?
- What diversion KPIs are public?
- Who audits overrides?
- Is surveillance ordinance review done?
- Value hypothesis
- Triaging AI is viable only with notice, override telemetry, and published safety metrics.
- Potential engagement
- Pre-procurement governance gap assessment against Seattle failure modes.
- Unsupported claims
- Human final authority alone equals safe deployment
- Guaranteed better patient outcomes
Pre-sales engineering
Role takeaway
- Fit
- Real-time 911 audio AI prompting with mandatory disclosure and audit logging.
- Architecture and integration
- Audio stream → AI prompts → dispatcher UI → override log → diversion/EMS outcome store.
- Prerequisites
- Legal/privacy clearance
- Notice mechanism
- Metrics pipeline
- Vendor training ban
- Constraints
- Multi-contractor nurse line
- Sensitive audio retention
- Public records exposure
- Security
- Encrypt audio features in transit/rest; minimize raw audio retention; ban secondary training.
- Proof of value
- Public diversion outcome report
- Override rate monitoring
- Completed surveillance review
Delivery
Role takeaway
- Work
- Implement notice, logging, KPI publication, and dispatcher policy for AI prompts.
- Dependencies
- Privacy counsel
- CAD/vendor
- Medical director
- Communications
- Ownership
- Fire EMS medical director owns clinical policy; city privacy owns disclosure compliance.
- Skills and adoption
- Train dispatchers to document disagreements with AI prompts.
- Governance checkpoints
- Surveillance/PIA approval
- Public metrics launch
- Quarterly outcome review
- Acceptance criteria
- Caller notice live
- Override logs complete
- First public metrics posted
- Risks
- Undisclosed listening
- Unmeasured harm
- Contract opacity
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Live-call AI listeners need explicit disclosure UX, immutable prompt/override logs, diversion outcome telemetry, and separation from vendor training uses.
Governance
Who approves, reviews and stays accountable for outcomes?
AI that influences 911 medical diversion should clear surveillance/privacy review before go-live and publish effectiveness metrics; post-hoc promises of openness are insufficient.
Security and privacy
What data, permissions and controls need testing?
Continuous AI listening to 911 medical calls is sensitive audio processing; require DPIA/surveillance ordinance paths, retention limits, and contractual bans on secondary training.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Undisclosed AI on medical 911 calls can especially affect limited-English and medically complex callers who cannot contest invisible triage.
Procurement
What should contracts, pricing and exit terms secure?
Require contract clauses for disclosure support, metric exports, training bans, and public pricing before live-call AI.
Operating model
Which teams own the service once it runs?
Live 911 AI prompting is an operating-model change needing public accountability, not a silent QA add-on.
Publication history
- 2026-10-02Emergency Services · Issue 295 resources
Stable resource ID: seattletimes-corti-911-ai-disclosure-20260614