Wallarm Leads Development of New A2AS Standard for AI Security

Wallarm has announced its leadership in the development of the "A2AS: Agentic AI Runtime Security and Self-Defense" framework, announced in a press release. This initiative, led by Eugene Neelou of OWASP and Wallarm, involves collaboration with researchers from AWS, Bytedance, Cisco, Elastic, Google, JPMorganChase, Meta, and Salesforce.

The A2AS framework introduces a new security layer for AI agents and LLM-powered applications, addressing risks such as prompt injection and agent compromise. It features three key capabilities: Behavior Certificates, Model Self-Defense Reasoning, and Prompt-Level Security Controls. These innovations aim to secure AI interactions and protect enterprises from potential security breaches.

Eugene Neelou, Head of AI Security at Wallarm, emphasized the importance of embedding security directly into AI agent runtimes. Ivan Novikov, CEO of Wallarm, highlighted the practical defense layer that A2AS provides, transforming theoretical security measures into actionable solutions. The A2AS paper is the first in a series aimed at establishing this framework as the industry standard for AI runtime security.