Last updated:
The operational, legal, and reputational danger arising from the use of incorrect, biased, or harmful content generated by an AI system.
While traditional IT focuses on protecting data inputs, AI governance must also manage outputs. Output risk includes hallucinations (plausible-sounding lies), algorithmic bias that leads to discriminatory decisions, and the inadvertent creation of content that infringes on third-party copyrights or trademarks.
Real world example:
A marketing team uses an unsanctioned AI to generate social media copy. The AI hallucinates a non-existent 50% discount - leading to a surge of angry customers and a legal requirement to honor the price.
