Agentic AI Runtime Security and Self-Defense (2025)
The A2AS framework has been introduced as a security layer for AI agents and applications. It aims to enforce certified behavior and ensure context window integrity while avoiding operational complexities. This paper outlines the BASIC security model that serves as the foundation for the A2AS framework.
- ▪The A2AS framework is designed to secure AI agents and LLM-powered applications, similar to HTTPS for HTTP.
- ▪It defines security boundaries, authenticates prompts, and applies custom policies to control agentic behavior.
- ▪The BASIC security model includes behavior certificates, authenticated prompts, security boundaries, in-context defenses, and codified policies.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Cryptography and Security arXiv:2510.13825 (cs) [Submitted on 8 Oct 2025] Title:A2AS: Agentic AI Runtime Security and Self-Defense Authors:Eugene Neelou, Ivan Novikov, Max Moroz, Om Narayan, Tiffany Saade, Mika Ayenson, Ilya Kabanov, Jen Ozmen, Edward Lee, Vineeth Sai Narajala, Emmanuel Guilherme Junior, Ken Huang, Huseyin Gulsin, Jason Ross, Marat Vyshegorodtsev, Adelin Travers, Idan Habler, Rahul Jadav View a PDF of the paper titled A2AS: Agentic AI Runtime Security and Self-Defense, by Eugene Neelou and 17 other authors View PDF Abstract:The A2AS framework is introduced as a security layer for AI agents and LLM-powered applications, similar to how HTTPS secures HTTP. A2AS enforces certified behavior, activates model self-defense, and ensures context window integrity.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv.org.
Discussion
0 commentsMore from arXiv.org
