Glossary term
Glossary term
Security
The protection of AI systems, models, data, prompts, tools, integrations, and outputs from compromise, misuse, leakage, manipulation, or disruption. AI security should be integrated into the secure development lifecycle and security operations, not treated as a separate research topic owned only by data science.
NIST SP 800-218A (May 2024) extends the Secure Software Development Framework to generative AI development, including model provenance, evaluation, and deployment security.
MITRE ATLAS is an adversary tactics, techniques, and procedures knowledge base for AI systems, modeled on the ATT&CK framework and updated quarterly.
The Cloud Security Alliance AI Controls Matrix (2024) provides 243 controls across responsible AI principles applicable to security teams.