What question did this study set out to answer?

To advance the security framework for agentic AI through the AFuller F7-LAS model.

January 20, 20260 citationsOpen Access

Securing Agentic AI: The AFuller F7-LAS™ (7-Layer) Model - v3.0

Key Points

To advance the security framework for agentic AI through the AFuller F7-LAS model.
Compared F7-LAS™ with existing frameworks like OWASP and NIST.
Introduced multi-agent privilege escalation scenario.
Developed governance crosswalk tables for NIST AI RMF and ISO 42001.
Removed complex academic formalism for a practitioner-friendly approach.
Eliminated protocol specificity to strengthen agnostic positioning.
F7-LAS™ demonstrates improved alignment with governance standards.
Multi-agent scenarios highlight potential security vulnerabilities.
The streamlined approach enhances accessibility for practitioners.

Abstract

What's New in Fuller 7-Layer Agentic AI Security (F7-LAS™) v3.0 Positioned F7-LAS™ against major frameworks (OWASP, NIST, MITRE, CSA) Added detailed multi-agent privilege escalation scenario Provided governance crosswalk tables for NIST AI RMF and ISO 42001 Removed academic POMDP formalism for practitioner focus Removed protocol-specific (MCP) to strengthen agnostic positioning.

اسأل الذكاء الاصطناعي

Bookmark

View Full Paper

Cite This Study

Fuller Anthony (Thu,) studied this question.

synapsesocial.com/papers/696f1a9f9e64f732b51eed7c https://doi.org/https://doi.org/10.5281/zenodo.18292122

اسأل الذكاء الاصطناعي

Bookmark

View Full Paper