From the Local Government edition of September 23, 2026
Colorado Springs goes live with Axon S.A.R.A.H. AI for non-emergency calls
Colorado Springs Police Department / City of Colorado Springs · Municipal public-safety communications and non-emergency resident access · Colorado Springs, Colorado, United States
- Publisher
- The Gazette
- Original publication
- September 23, 2026 The Gazette reporting on Tuesday go-live
- Source retrieved
- 2026-09-24
- Event date
- 2026-09-23
What happened
On Sept. 23, 2026 The Gazette reported that an AI chatbot named S.A.R.A.H. (Service Assessment and Response Assistance Helper), powered by Axon’s Prepared AI, began answering all non-emergency calls to the Colorado Springs communications center that Tuesday. Deputy Police Chief John Koch said the goal is shorter waits so dispatchers can prioritize 911: emergency answer times average 11–13 seconds while non-emergency waits usually exceed 30 seconds; with AI there is now almost no wait for non-emergency calls. Across four beta tests, CSPD found roughly half of non-emergency calls could be completed without a human dispatcher; the bot asks questions to decide whether a human call taker is needed. Koch said humans review each AI-produced call and callers can request a human. Contract cost for Prepared AI by Axon is $170,000 annually per a CSPD spokesperson. CSPD already uses other Axon tools (drones, body-cam report drafting, LPR). Koch discussed possible future developmental partnering including remote dispatcher answering.
Why it matters
Same-day U.S. city go-live of vendor non-emergency voice AI with published beta completion share, wait-time claims, annual contract cost, and mandatory human review—directly relevant to municipal resident-access and communications capacity.
Evidence and measured results
Gazette Sept. 23, 2026: S.A.R.A.H. live Tuesday for all non-emergency calls; Axon Prepared AI; $170,000 annually; four beta tests ~half completed without human dispatcher; human review of each AI-produced call; caller can request human; emergency 11–13s vs non-emergency >30s baseline; almost no wait after go-live per police.
Limitations and uncertainty
Local press quoting CSPD; beta “roughly half” and wait-time claims are department-reported, not an independent audit; Axon ecosystem concentration noted in article but not evaluated.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-09-24; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Non-emergency callers wait while 911 demand competes for the same communications-center capacity.
- Stakeholders
- Police/communications leadership
- City manager/council
- Dispatchers/call takers
- Residents
- Procurement/legal
- Discovery questions
- What share of non-emergency calls completed without humans in local betas?
- Who reviews every AI-produced call and within what SLA?
- How is the 911 path technically and operationally separated?
- What is the annual contract cost and vendor concentration risk?
- Value hypothesis
- Voice triage can cut non-emergency wait if beta completion rates hold and humans remain in every decision path—without proving audited FTE savings yet.
- Potential engagement
- Time-boxed non-emergency AI triage design with human-review SOP, metric dashboard, and privacy controls—no guaranteed half-call automation.
- Unsupported claims
- Guaranteed elimination of non-emergency wait
- Independent audit of the half-call beta result
- Replacement of human dispatchers
- Safe extension to 911 without separate authorization
Pre-sales engineering
Role takeaway
- Fit
- Vendor voice AI front-ending municipal non-emergency call taking with mandatory human review.
- Architecture and integration
- PSTN/non-emergency line → Prepared AI triage → human review queue / optional human request → CAD/records actions as today.
- Prerequisites
- Communications-center telephony integration
- Human review staffing model
- Retention/privacy policy
- Caller disclosure script
- Constraints
- Department-reported metrics only
- Axon suite coupling
- Not for 911 automation in this report
- Local press not an audit
- Security
- Encrypt and tightly ACL call audio/transcripts; ban secondary model training on resident calls without council-approved terms.
- Proof of value
- Wait-time before/after
- Share completed without human
- Human-override rate
- Complaint/escalation rate
Delivery
Role takeaway
- Work
- Configure triage flows, train reviewers, publish resident notice, and run a 30-day quality sample with weekly fixes.
- Dependencies
- CSPD/comms center
- Axon Prepared AI
- City procurement/legal
- Public information office
- Ownership
- Police/communications owns operations; city IT/procurement owns vendor controls.
- Skills and adoption
- Teach call takers to challenge AI triage and document overrides; brief council on metrics.
- Governance checkpoints
- Pre-go-live human-review SOP
- 30-day metric review
- Annual contract/privacy review
- 911-separation attestation
- Acceptance criteria
- Human reviews each AI-produced call
- Caller can request a human
- 911 path unchanged
- Published cost and beta basis
- Risks
- Silent wrong triage
- Vendor lock-in across Axon suite
- Privacy of voice data
- Over-claiming wait-time gains
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?
Non-emergency voice triage sits in front of the communications center with explicit human review and human-request fallback; separate from 911 priority path.
Governance
Who approves, reviews and stays accountable for outcomes?
Publish beta metrics, annual cost, and human-review rules at go-live; treat Axon suite concentration as a vendor-concentration risk for council oversight.
Security and privacy
What data, permissions and controls need testing?
Voice transcripts and triage decisions are sensitive; require retention, access, and misuse controls beyond marketing confidence claims.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Near-zero non-emergency wait can improve access if language and disability paths remain, and if human request is easy for complex or emotional cases.
Procurement
What should contracts, pricing and exit terms secure?
$170k annual Prepared AI should be tied to measurable review SLAs, data-use limits, and exit provisions given existing Axon footprint.
Operating model
Which teams own the service once it runs?
Keep humans as reviewers and fallback; use AI to absorb routine non-emergency volume so 911 priority stays protected.
Publication history
- 2026-09-23Local Government · Issue 184 resources
Stable resource ID: colorado-springs-sarah-axon-nonemergency-20260923