What question did this study set out to answer?

The aim is to map the OWASP Agentic AI risks to relevant decision-making controls within AgentOS.

June 3, 2026Open Access

OWASP Agentic AI Top 10 -- Decision Evidence Mapping

Key Points

The aim is to map the OWASP Agentic AI risks to relevant decision-making controls within AgentOS.
Mapped 10 OWASP Agentic AI risks to specific Decision Passport controls.
Distinguished between pre-execution prevention and post-execution recording for each risk.
Ensured regulatory alignment with multiple key frameworks and acts.
Established clear mapping for all 10 risks to enhance decision-making.
Highlighted compliance with OCC Bulletin 2026-13 and related regulations.
Provided a structured approach to managing AI risk within the outlined frameworks.

Abstract

Version 1.1 — corrects signing substrate framing (Risk 3 and Differentiation table) and OCC Bulletin 2026-13 status. See Version History section in the document.---Maps all 10 OWASP Agentic AI risks to AgentOS Decision Passport controls. Distinguishes pre-execution prevention from post-execution recording per risk. Covers regulatory alignment with OCC Bulletin 2026-13, NYDFS Part 500, EU AI Act Article 12, NY RAISE Act, SOC2 CC6.1/CC4.1, and GDPR Article 22. Published by AgentOS Technologies, Inc. Citable reference: https://doi.org/10.5281/zenodo.20486369

OWASP Agentic AI Top 10 -- Decision Evidence Mapping

Key Points

Abstract

Cite This Study