Joshua Beck
Application Security Architect

Specializing in AI Security Innovation & Application Security.

OWASP Agentic AI Security

April 2025 - Present

OWASP GenAI Security Project

I support the OWASP GenAI Security Project through contributions to multiple agentic AI documents, including contributing as an Entry Lead for the OWASP Top 10 for Agentic Applications 2026. This flagship offering from OWASP provides security professionals across the world with vital resources to support the security of their AI applications. My contributions not only add to the incredible wealth of knowledge from the most accomplished security professionals in the AI space, those seasoned experts also give me a feedback loop to further develop my agentic AI security skills.

My Contributions

  • OWASP Top 10 for Agentic Applications 2026: The pinnacle of OWASP's Agentic Security Initiative! The top 10 will be an incredible resource for every developer looking to secure notorious and complex agentic systems. I'm so proud to have had a role in the development of this list, and I am excited to see where it takes us in the years to come.
  • Co-lead for the MCP CheatSheet: This cheat sheet provides desparately needed industry advice for consuming local and remote MCP servers.
  • Securing Agentic Applications Guide: This guide is designed to provide in-depth technical guidance to any agentic AI developer. From architecture to specific authentication or logging controls, we aim to provide a complete how-to guide for securing agentic applications.
  • State of Agentic AI Security and Governance 1.0: Every C-suite level security professional needs to have high-quality guidance. Our goal with the State of Agentic AI Security document was to provide that guidance, spreading necesary data and informing decision makers on critical and timely AI security information.
  • AI Vulnerability Scoring System (AIVSS): As a founding member of the AI Vulnerability Scoring System, I not only contributed valuable suggestions and edits to this ground breaking approach to AI vulnerabilities and risk management.