AI Incidents

Track AI-specific incidents from initial report through resolution. Manage response timelines, coordinate tasks, and handle Art. 62 authority notifications for serious incidents involving high-risk AI systems.

Incident Types

AI SENTINEL tracks incident types specific to AI systems, distinct from data protection incidents handled in DPO Central.

Hallucination
Bias / Discrimination
Model Drift
Adversarial Attack
Prompt Injection
Unauthorized Access
Safety Failure
Performance Degradation
Data Poisoning
Privacy Violation

Incident Lifecycle

Every incident follows a structured workflow. Status transitions and timestamps are tracked for compliance evidence.

Reported
Investigating
Mitigating
Resolved
Closed

Severity Classification

Severity levels prioritize response efforts and determine notification requirements.

Low

Minor incident, limited impact. No regulatory notification required.

Medium

Moderate impact, contained quickly. Internal review recommended.

High

Significant impact. Likely requires authority notification under Art. 62.

Critical

Large-scale failure or safety risk. Immediate Art. 62 notification required.

Art. 62 Authority Notifications

Under EU AI Act Article 62, providers of high-risk AI systems must report serious incidents to the relevant market surveillance authority. AI SENTINEL tracks notification deadlines, recipient authorities, and submission status.

When to notify

Serious incidents involving death, serious damage to health, property, or the environment, or serious and irreversible disruption of critical infrastructure.

Timeline

Report to the authority without undue delay after becoming aware of the incident. The system tracks elapsed time from the incident report date.

What to include

Description of the incident, affected AI system, severity, root cause analysis (if available), and corrective measures taken or planned.

Incident Timeline

Each incident maintains a detailed timeline recording all events, status changes, task assignments, and communications.

09:15CREATEDIncident reported
09:45INVESTIGATINGAssigned to investigation team
11:00UPDATERoot cause identified — model drift detected
14:00NOTIFICATIONArt. 62 notification sent to authority
16:30MITIGATINGMitigation deployed — model rolled back
18:00RESOLVEDIncident resolved and closed

Reporting an Incident

1

ReporterReport the incident

Navigate to Operations → Incidents and click Report Incident. Provide initial details: affected AI system, incident type, and description.

2

AI OfficerTriage and assign severity

Review the report, assign a severity level, and determine the response scope.

3

IT TeamInvestigate and contain

Gather evidence, identify the root cause, and implement immediate containment measures.

4

AI OfficerNotify authorities if required

For high-risk systems with serious incidents, prepare and submit the Art. 62 notification.

5

AI OfficerResolve and close

Document the resolution, lessons learned, and any corrective actions. Close the incident.