An Ethical Impact Assessment (EIA) is a structured review of an AI system or deployment to identify risks, benefits, and mitigations before and during use. It typically examines privacy, bias, safety, misuse potential, human oversight, and the controls needed to reduce harm. In practice, an EIA turns abstract principles into testable questions: who can access the model, what data it uses, how outputs are reviewed, and what happens if the system behaves unexpectedly.
In cyber security, EIAs matter because AI can expand attack surface and operational risk. A poorly assessed chatbot, classifier, or decision engine may leak sensitive data, enable social engineering, or make unsafe automated decisions. Defenders use EIAs to document safeguards such as access controls, logging, rate limits, red-teaming, monitoring for model drift, and rollback plans. That makes EIA useful not only for compliance, but also for proving that security and governance controls work in real deployments.