TLDR: Token Security and Descope, in partnership with prominent cybersecurity leaders, have launched the ‘AI Security Guide: A Maturity Model for Secure Agentic AI Adoption.’ This new framework addresses the evolving security challenges posed by autonomous AI systems, providing organizations with a roadmap for responsible and scalable AI deployment.
LOS ALTOS, Calif. and NEW YORK – August 11, 2025 – As agentic AI rapidly reshapes enterprise operations, a critical new frontier in cybersecurity has emerged. In response, Token Security and Descope, collaborating with a distinguished group of cybersecurity leaders, have announced the release of the ‘AI Security Guide: A Maturity Model for Secure Agentic AI Adoption.’ This comprehensive framework is designed to empower organizations to integrate agentic AI securely, responsibly, and at scale, while simultaneously fostering innovation.
The guide synthesizes practical insights and best practices from across the security community, featuring contributions from CISOs at leading organizations including Silicon Valley CISO Investments (SVCI), Vercel, Verily Life Sciences, Live Oak Bank, AppLovin, Notable Capital, and Xcel Energy.
Itamar Apelblat, Co-Founder and CEO of Token Security, emphasized the paradigm shift: “We’re entering a world where AI doesn’t just suggest, it acts. Agentic AI systems are launching code, triggering workflows, making decisions, and creating new identities. As a result, Non-Human Identities will quickly outnumber human identities by more than 100 to one. Security must now shift priority to the identities that matter most in this new world.” Token Security specializes in Non-Human Identity Security, with its NHI platform designed to mitigate risks and accelerate AI adoption by providing tools for discovering, governing, securing, and threat detection for NHIs.
Rishi Bhargava, Co-Founder of Descope, highlighted the impetus behind the guide: “We created this guide because we kept hearing the same question from CISOs and developers; ‘How do I make sure our AI or MCP server doesn’t break things or worse?’ The answer starts with treating AI agents like first-class actors. Secure authentication, granular authorization, and policy-based governance needs to be baked into every agentic AI deployment. We’re proud to team up with the best minds in the industry to offer a roadmap to ensuring secure AI and MCP deployments.”
The ‘AI Security Guide’ introduces a four-phase maturity model specifically designed to help enterprises navigate and mitigate the risks associated with AI autonomy. This model spans from initial ‘shadow AI’ experimentation to the secure, full-scale deployment of autonomous agents operating across critical systems.
Also Read:
- Fortifying Critical Infrastructure Against Autonomous AI Threats
- NIST Embeds AI Cybersecurity Risks into Existing Frameworks, Opting Against New Mandates
Shaun Marion, VP and CISO at Xcel Energy, underscored the guide’s timely relevance: “This guide arrives at a critical moment to provide security leaders a playbook to ensure we don’t move into a future shaped by invisible AI actors, untraceable actions, and preventable failures.” The collaborative effort aims to provide the necessary tools and knowledge for organizations to confidently embrace the transformative potential of agentic AI while maintaining robust security postures.
