---
title: "AI incident summaries — Lemwatch Help"
description: "Plain-English summaries of every incident — what happened, what changed, and the likely cause — generated the moment an alert fires."
canonical: "https://lemwatch.com/help/ai-incident-summaries"
source: "https://lemwatch.com"
---

# AI incident summaries

> Plain-English summaries of every incident — what happened, what changed, and the likely cause — generated the moment an alert fires.

## Quick answer

 Every incident gets a 2–3 sentence AI summary pinned to the top: what triggered, which check, which region, and the closest recent change (deploy, DNS, cert).

## What the summary includes

- The specific check that failed (HTTP 5xx, SSL expiry, DNS mismatch…)

- Regions and time window

- Recent changes correlated within 30 minutes

- Suggested next step (link to a troubleshooting article)

## Where to find it in Lemwatch

 **Incident detail → Summary.** Also posted to Slack / email alerts when integration is on.

## How to trust it

- Every claim links to underlying evidence (raw check logs, DNS diff, deploy hook)

- Confidence badge shows Low / Medium / High

- Low-confidence summaries are marked "unverified — investigate manually"

## FAQ

### Which model powers it?

 Lemwatch AI Gateway — model choice is abstracted; upgrades happen server-side.

### Does it count against my AI quota?

 Yes — see [AI usage and quotas](https://lemwatch.com/help/ai-usage-and-quotas).

## Related

- AI root-cause analysis

- Autopilot overview

- Alert transparency and evidence


## Related

- [Help centre](https://lemwatch.com/help)
- [AI root-cause analysis](https://lemwatch.com/help/ai-root-cause-analysis)
- [What is Autopilot in Lemwatch?](https://lemwatch.com/help/autopilot-overview)
- [Alert confirmation gates](https://lemwatch.com/help/alert-confirmation-gates)
- [AI alert noise reduction](https://lemwatch.com/help/ai-alert-noise-reduction)
- [Lemwatch pricing](https://lemwatch.com/pricing)