Monitor
AI incidents
Alerts when visibility drops, mentions disappear, citations shift, or narratives change.
What is AI incidents?
AI incidents automatically detect meaningful changes between prompt runs and open alerts with severity, description, and root cause analysis—including contributing factors and recommended actions.
How does it work?
- 1Complete a new prompt run after a baseline run exists.
- 2Obsurfable compares each question's mention, citation, entities, and response text.
- 3Incidents open for visibility drops, lost mentions, narrative shifts, and more.
- 4Review root cause analysis on the Incidents page; acknowledge or resolve as you act.
- 5Optional email alerts in Settings when incidents are detected.
Why it matters
Rankings and scores alone do not tell you when something broke. Incidents tell you what changed, when, and what to do next.
Related features
- AI Brand Health — Datadog-style overview — health score, open incidents, platform status, and trend.
- Prompts — Ask ChatGPT with web search on buyer questions; track mentions and citations run over run.
- Entity perception — What does AI think you are? Entity ownership, drift, and influence pathways.
- AI Visibility Score — One score for mentions, citations, entity listings, competitor gaps, and prioritized fixes.
Ready to try AI incidents?
View plans