Thoughts on AI Security Architecture & Agentic Security
All the articles I have published.
-
Visualizing the Invisible: Why Agentic AI Needs a New Approach to Security Understanding
Published date:9 min readIntroducing Agentic Security Explorer (ASE) - As agentic systems grow more complex, traditional approaches struggle to explain how components, threats, and business impacts connect. Learn how the Agentic Security Explorer (ASE) helps visualize the complete agentic attack surface.
-
Securing Agentic AI with Trust Collapse Indicators (TCIs)
Published date:8 min readTrust Collapse Indicators (TCIs)- A New Lens for Securing Agentic AI Systems. Trust Collapse Indicators (TCIs) provide a forward-looking way to identify weakening trust relationships before they become incidents or full business impact across complex agentic and multi-agent systems.
-
Agentic Security - The Rise of the Out-of-Bound Agent
Published date:6 min readSATF Series- Why autonomous systems need continuous trust boundaries. The Rise of the Out-of-Bound Agent
-
Inside Secure Agent Trust Framework (SATF)
Published date:5 min readSATF Series- Part 4- Agent Trust Fabric, Delegation Provenance, Goal Integrity, and Continuous Reassessment.
-
Introducing the Secure Agent Trust Framework (SATF)
Published date:5 min readSATF Series- Part 3- End-to-End Enterprise Framework for Autonomous Agent Governance and Contextual Security, trusted autonomous actions and secure agentic outputs.
-
The Agent Trust Is Not Granted Once - Secure Agent Trust Framework (SATF)
Published date:4 min readSATF Series- Part 2- Why agent trust requires more than IAM and static permissions.
-
The Agent Trust Gap - Secure Agent Trust Framework (SATF)
Published date:5 min readSATF Series- Part 1- Why recent agent incidents show standalone security models are necessary but not sufficient.
-
Contextual Agent Security: Is Agent Authorization the Next IAM?
Published date:9 min readGoogle Research''s Conseca framework may signal a fundamental shift from permission-based security toward context-aware authorization for autonomous AI agents.
-
Fable 5 (Mythos)- From AI Guardrails to an AI Cyber Control Plane
Published date:6 min readSWOT analysis of Anthropic Fable 5 cyber safeguards, CJS jailbreak framework, and the shift from AI guardrails to an AI cyber control plane
-
Top 10 Priorities for a New Product Security Leader in a Global Healthcare Company
Published date:11 min readA secure-by-design, quality/Safety first leadership playbook for healthcare manufacturing product security across legacy, connected, AI-enabled, regulated, and non-regulated products.
-
Next-Gen Zero Trust Security Architecture for AI- A 2026+ Enterprise Reference Model
Published date:24 min readA 2026+ Enterprise Reference Architecture View for Securing GenAI, RAG, Multi-Agent Systems, and Traditional ML Pipelines. A CISO-oriented guide to securing the full AI lifecycle across GenAI, RAG, agentic AI, multi-agent systems, and traditional ML using Zero Trust, threat modeling, assurance, and AI SecOps.
-
Navigating Supply Chain Risks and Mitigations in a Layered AI SecOps Architecture
Published date:6 min readSecuring the AI Frontier-Navigating Supply Chain Risks and Mitigations in a Layered AI SecOps Architecture Framework across four horizons =>