Story perspectives

OWASP Updates AI Security Guidance, Identifies 21 Key Risks

4/6/2026

Summary

Story summary

The Open Web Application Security Project (OWASP) updates AI security guidance, with guides for generative AI and agentic AI identifying 21 risks, including sensitive data leakage and unauthorized data flows.
OWASP will switch to a six-month update cycle.
Arkose Labs finds 97% of security leaders expect AI-driven incidents within a year, but only 6% of budgets address risk.
Organizations must improve visibility and governance for AI agents to mitigate threats.

History

Historical Context

The General Data Protection Regulation (GDPR), implemented by the European Union in 2017, established stringent data protection standards, influencing organizations like OWASP to enhance security measures. As AI-related incidents become a growing concern, OWASP's recent update identifies 21 risks associated with AI, underscoring the ongoing need for robust governance in technology and the lasting impact of GDPR on data privacy practices.

1 of 2

Story summary

The Open Web Application Security Project (OWASP) updates AI security guidance, with guides for generative AI and agentic AI identifying 21 risks, including sensitive data leakage and unauthorized data flows.
OWASP will switch to a six-month update cycle.
Arkose Labs finds 97% of security leaders expect AI-driven incidents within a year, but only 6% of budgets address risk.
Organizations must improve visibility and governance for AI agents to mitigate threats.

1 / 2