Glossary · AI governance
AI Risk
The risks specific to artificial intelligence systems, including bias, opacity, unsafe outputs, and loss of human control.
Put the term into practice: KaitoSec AI
AI risk covers the harms that can arise from developing or using AI, beyond ordinary IT risk. These include biased or discriminatory outcomes, lack of transparency, unreliable or unsafe outputs, privacy impacts, and the erosion of meaningful human oversight.
An AIMS under ISO/IEC 42001 manages these risks across the AI lifecycle, and the EU AI Act ties obligations to the level of risk a system poses.
