Since December 2023 the Seattle Fire Department (SFD) deployed Corti, an artificial intelligence developed by a Copenhagen startup called Corti, on every medical 911 call. The system analyzed calls in real time and provided prompts to dispatchers; in a number of cases those prompts led dispatchers to redirect callers to a nurse advice line based in Texas.
Purpose and operational detail
The stated rationale was pragmatic: diverting low‑acuity calls to nurse lines can free ambulances and paramedics for true emergencies. In practice, however, SFD ran Corti on all medical calls and used its output to influence whether to send emergency resources or instead refer the caller to the Texas nurse line.
Lack of disclosure and approvals
The program was not disclosed to the public. According to reporting, the Seattle Fire Department did not register the system under the city’s surveillance ordinance and did not seek approval from the Seattle City Council before putting the tool into operation. The deployment remained undisclosed for over two years and was revealed this week by GeekWire.
Impact and risks
Reports indicate a roughly 32% increase in diversions after Corti was introduced — meaning about one in three more callers were steered away from ambulance dispatch than before. There has been no public audit released detailing who was affected by these diversions or whether any adverse outcomes occurred. The central concern is that automation was applied at a decision point where an incorrect recommendation can have life‑threatening consequences.
Why this matters
Reducing dispatcher workload and appropriately triaging calls are legitimate and important objectives. The controversy here centers on process and oversight: SFD applied an automated judgment tool to critical, potentially life‑saving triage decisions while bypassing mandated disclosure, public review and city oversight mechanisms designed to catch and mitigate risks.
Next steps and unresolved questions
Following the reporting, local officials and oversight bodies are likely to seek explanations and may open reviews or investigations. Key unresolved questions include whether any serious adverse events resulted from diversions, what data and criteria Corti used to make recommendations, and what safeguards will be required before similar technologies are deployed in public safety contexts in the future.
Summary
Seattle’s deployment of Corti highlights the tension between operational efficiency and public accountability: a fire department adopted an AI to improve resource allocation, but did so without public notice or regulatory approvals, raising concerns about safety and oversight when automation affects life‑and‑death dispatch choices.
