Unit 6: Incident Reporting – Know What to Flag and How
An AI incident is any situation where an AI system causes — or nearly causes — harm, failure, or misuse.
Examples include:
Bias or discrimination (e.g., unfair hiring decisions)
Misinformation (e.g., incorrect medical or legal advice)
Privacy violations (e.g., sharing sensitive data without consent)
System errors (e.g., inaccurate outputs in high-stakes contexts)
Loss of control (e.g., AI acting outside of expected parameters)
📌 If the system impacts real people and something goes wrong — it may be an incident.
