Skip to main content

What is alert triage?

Alert triage is the first review of a security alert: confirm what happened, judge severity, and close, escalate, or respond.

Glossary
Updated 2 min read

Definition

Alert triage

Also called Security alert triage, SOC triage, Incident triage

Alert triage is the first review of a security alert, in which an analyst or automated system confirms what happened, judges whether it is malicious and how severe it is, and decides to close it, escalate it, or start a response.

How it works

NIST SP 800-61r3 describes triage as a preliminary review that verifies a cybersecurity incident has occurred, then estimates its severity and the urgency of response. Before that point, it recommends tools that filter large volumes of potentially adverse events down to what humans can review, and integrating threat intelligence and other context into the analysis.

For a single alert, triage usually means answering a short list of questions. What fired and why? Who is the user or what is the host, and is this normal for them? What else happened around the same time? Does threat intelligence know the indicator? The analyst then records a disposition, such as true positive, benign true positive, or false positive, and either closes the alert, escalates it into an incident, or starts a response from a playbook.

When several alerts compete for attention, NIST suggests prioritizing incidents by their scope, likely impact, time-critical nature, and resource availability. A simple working version:

FactorRaises priority
AssetDomain controllers, identity providers, production data, executives' accounts
Evidence of successA successful sign-in, process execution, or data transfer, not just an attempt
SpreadThe same indicator or behavior on more than one host or account
Time pressureActive sessions, ongoing exfiltration, or ransomware staging

Two habits make triage compound over time: write down why each alert was closed, and send every false positive back to whoever owns the detection, so the same noise doesn't return tomorrow.

Example

An alert reports repeated denied MFA pushes for one user. Triage checks the identity logs: the denials came from one enrolled device on a hotel network, with no successful sign-in from an unfamiliar location, and the user confirms they were traveling and fumbling the prompt. The alert is closed as a false positive with a note. Cotool's response page shows an agent reaching that same disposition for that same scenario.

Why it matters for a SOC

Triage is the SOC's highest-volume task and the step where noise turns into alert fatigue, so it largely determines time to respond, one of the SOC metrics most teams track. That is why it is the first job handed to automation and AI agents. In Cotool's AI SOC, agents pick up every alert as it lands, gather context across the connected tools, and close or escalate it with the evidence attached, while the team controls the instructions they follow and approves state-changing actions.

Go deeper

Sources

  1. NIST SP 800-61 Rev. 3: Incident Response Recommendations and Considerations for Cybersecurity Risk Management · accessed November 16, 2026
  2. AI SOC Without the Black Box | Cotool · accessed November 16, 2026
  3. AI Alert Triage and Incident Response Agents | Cotool · accessed November 16, 2026

Attackers are scaling with tokens

Cotool helps defenders operate at machine speed. See how security teams are scaling Detection & Response.

Book a demo