Cases › operational mistake › LR-2026-0298

Ava AI agent misclassified emergency caller as nonemergency during 2024 rollout

Preliminary Why: Publisher not yet named · subject to change

Severity minor: The misclassified call was quickly transferred to a human dispatcher, contained to one call with no third party harmed.

Earliest source we hold 2026-10-31 · Captured by AI Risk Wire 2026-10-31 20:30 UTC

Ava, an AI agent made by Seattle startup Aurelian, answers nonemergency calls for Snohomish County 911 and other agencies.

Risks this incident falls under

Risk	Role	NIST AI 600-1	Incidents	Last 6 quarters
Agent takes destructive or unapproved actions	Primary	Human-AI Configuration	18

Sorted by rule from the case's fields; not yet read from the sources.

Control failure chain

L1 · Model Not read yetAurelian

L2 · Harness Not read yetAva

L3 · Application Not read yetDeployer not named

L4 · Infrastructure / target Not read yetOrganisation hit

L1 · ModelAurelianModel provider

Sources not yet read for controls.

Model version not linked to a model card.

L2 · HarnessAvaAgent framework vendor

Sources not yet read for controls.

Other controls for this risk at this layer, not yet read for: Tool allow-list, Fail-closed defaults, Stop control.

L3 · ApplicationDeployer not namedDeployer

Sources not yet read for controls.

Other controls for this risk at this layer, not yet read for: Human approval for high-impact actions, Change review and separation of duties, Least-privilege access for the agent, Production kept apart from test, Activity logs, Adversarial testing before deployment.

L4 · Infrastructure / targetOrganisation hit

Sources not yet read for controls.

Other controls for this risk at this layer, not yet read for: Backups and recovery.

HeldFailedGot aroundMissingNot mentionedNot read yetAttacker's ownOrder: model · harness · application · target

The same controls in other incidents of this risk

Control	Layer	Here	Held elsewhere	Did not hold elsewhere
Tool allow-list	L2 Harness	Not read yet	none	none
Stop control	L2 Harness	Not read yet	none	none
Fail-closed defaults	L2 Harness	Not read yet	none	none
Least-privilege access for the agent	L3 Application	Not read yet	none	none
Production kept apart from test	L3 Application	Not read yet	none	1
Human approval for high-impact actions	L3 Application	Not read yet	none	3
Change review and separation of duties	L3 Application	Not read yet	none	none
Activity logs	L3 Application	Not read yet	none	none
Adversarial testing before deployment	L3 Application	Not read yet	none	none
Backups and recovery	L4 Infrastructure / target	Not read yet	none	none

Remediation

Fix reported

“Updates to Ava since then factor in situations like this.”

No later incident for Aurelian under this risk on record.

Risk grouping: Not yet classified

No source we hold states: who caused it, intended or not, before or after release, the ai's part.

The story

1 · What it was doingAva, an AI agent built by Seattle startup Aurelian, answers nonemergency calls for Snohomish County 911 and other agencies, processing conversations, choosing responses and deciding when to transfer callers to human dispatchers.

2 · What went wrongAva treated an emergency as a nonemergency: a caller complaining of headaches should have been immediately flagged as an emergency but was not.

3 · What it reached or didThe call reached a human dispatcher after a quick transfer, so the caller was not left without help.

4 · What was doneThe call was transferred to a dispatcher, and Keenan said updates to Ava since then factor in situations like this.

No source states: how it was caught, when it happened, when it was found, how many records were exposed, the direct loss, a count of the other parties affected, any court case, the cause, what set it off.

Show every field, with its source and grade

The AI system

Developer

Aurelian

C one secondary source states it

Software around the model

Ava

C one secondary source states it

What it was allowed to do

Answering nonemergency call lines and transferring calls to human dispatchers

C one secondary source states it

How much it acted alone

it acted alone, watched by a person

D inferred by us from the described behaviour, reason shown

Stage of use

in live use

C one secondary source states it

Whose system

company using AI

D inferred by us from the described behaviour, reason shown

How it was connected

built into a product

D inferred by us from the described behaviour, reason shown

What it was for

Answering nonemergency calls and routing or transferring callers to human dispatchers

C one secondary source states it

What happened

Earliest source we hold

2026-10-31

A the date the source itself carries

Who was responsible

the model provider

D inferred by us from the described behaviour, reason shown

What was done

Updates to Ava since then factor in situations like this.

C one secondary source states it

What went wrong

operational mistake

D inferred by us from the described behaviour, reason shown

AI agent

an AI agent acted

C one secondary source states it

Impact

Severity

minor

D inferred by us from the described behaviour, reason shown

Kind of harm

operations

D inferred by us from the described behaviour, reason shown

Who was affected

A caller following up on an incident from two weeks earlier who was complaining of headaches

C one secondary source states it

Why this severity

The misclassified call was quickly transferred to a human dispatcher, contained to one call with no third party harmed.

D inferred by us from the described behaviour, reason shown

Context

Industry

Emergency call centers (911 dispatch)

C one secondary source states it

Country

United States

C one secondary source states it

Base count

192,000 units, 2026

D inferred by us from the described behaviour, reason shown

Shared exposure

one shared model or part put many users at risk

C one secondary source states it

LR-2026-0298 · version 2 · published 2026-10-31 after automatic checks; not yet checked field by field by a person · changes to this case · JSON