Glossary: Application and Credential Attacks, and AI Risks

integrity

Data or decisions being correct and unaltered. Bias harms the integrity of an AI's decisions because the output is wrong.