AI Incident Reporting

AI Incident Reporting

An AI system starts rejecting whole regions. The clock runs.

What Is AI Incident Reporting?

Article 73 of the EU AI Act requires providers of high-risk AI systems to report serious incidents to market surveillance authorities within strict deadlines, and deployers must inform the provider and the authority. In this exercise, you investigate an AI-powered customer onboarding system that has been rejecting applicants from specific regions, determine whether the anomaly meets the Article 73 threshold for a serious incident, contain the damage, file the report, and resist vendor pressure to delay.

What You'll Learn in AI Incident Reporting

AI Incident Reporting — Training Steps

  1. Article 73: Serious Incident Reporting

    When a high-risk AI system causes a serious incident, the EU AI Act sets strict reporting deadlines. Providers report it to the market surveillance authority (Article 73); deployers inform the provider and the authority (Article 26(5)). A serious incident means death or serious harm to health, serious and irreversible disruption of critical infrastructure, an infringement of EU law protecting fundamental rights, or serious harm to property or the environment (Article 3(49)). Today, one of your company's AI systems has triggered exactly this.

  2. An Urgent Support Escalation

    Alice receives an urgent email from the customer support team. They have flagged a disturbing pattern in the OnboardAI system - the company's AI-powered customer onboarding tool.

  3. Checking the AI Dashboard

    Alice opens the AI monitoring dashboard to confirm the anomaly with hard data. The dashboard tracks all AI system metrics including approval rates, alerts, and operational costs.

  4. Confirming the Anomaly

    The dashboard confirms what customer support reported. The critical alert shows a clear, systematic pattern of regional discrimination by the OnboardAI system.

  5. Is This a Reportable Incident?

  6. Immediate Containment

    Does this meet the Article 3(49) threshold for a 'serious incident'? Yes: an incident that directly or indirectly leads to an infringement of Union law protecting fundamental rights qualifies, and systematic discrimination against 230 applicants breaches the right to non-discrimination. Systematic pattern: 0% approval (100% rejection) sustained over 7 days Fundamental rights impact: non-discrimination is a protected right Significant scale: 230 people denied services based on their region Ongoing harm: the system is still rejecting applicants The first priority is containment. From the AI dashboard, Alice halts OnboardAI and routes applications from the three affected regions to manual processing. The correct order: contain first , then investigate and report. Never wait for root cause analysis before stopping ongoing harm.

  7. Signing in to the Incident Portal

    With the system contained, Alice opens the internal incident reporting portal. Access is restricted to authorized operations staff - she signs in with her corporate credentials before filing the report.

  8. Filing the Incident Report

    Alice fills in the incident report with the facts the support team confirmed and the containment actions the operations team has already taken. This report will route to the compliance team, who will forward it to the market surveillance authority.

  9. Why Each Section Matters

    The report is now part of the official record. Each section serves a specific purpose for the compliance team and the market surveillance authority.

  10. Reporting Timelines

    The report has been submitted internally. Now the legal clock starts. Under Article 73, providers of high-risk AI systems report serious incidents to the market surveillance authority: Immediately once a causal link, or its reasonable likelihood, is established No later than 15 days after the provider becomes aware No later than 2 days for a widespread infringement or serious, irreversible disruption of critical infrastructure; 10 days if a person has died Deployers inform the provider immediately, then the authority (Article 26(5)), and both cooperate with any investigation. SynapseAI deploys OnboardAI but did not build it, so it must notify the vendor immediately.

Security Framework Coverage

CIS Controls

  • CIS 17 Incident Response Management
  • CIS 14.6 Train Workforce Members on Recognizing and Reporting Security Incidents

NIST CSF

  • PR.AT-01 Personnel are provided with awareness and training so that they possess the knowledge and skills to perform general tasks with cybersecurity risks in mind
  • RS.MA Incident Management

NICE Framework

  • OG-WRL-008 Privacy Compliance
  • OG-WRL-002 Cybersecurity Policy and Planning

EU AI Act

  • Art. 73 Reporting of serious incidents