The perimeter moved
Security boundaries now form around agent knowledge, authority, tools, runtime, outputs, and leakage paths.
IMS
The New Perimeter
Securing the Agentic Enterprise
The book behind the framework, the practice, and the seven AI attack surfaces.
Flagship brand asset
The book for leaders responsible for AI adoption without catastrophic risk.
The New Perimeter gives CISOs, CIOs, CAIOs, boards, and technical leaders a shared language for securing agents, tools, context, identity, and egress.
Central argument
The book makes one idea unavoidable.
Agents are not just software features. They are delegated actors. Enterprise security must shift from securing places to securing agent behavior.
Governance must be operational
Policies that do not map to controls, owners, tests, and evidence will not survive agentic scale.
Speed requires discipline
AI transformation without security architecture is not speed. It is unmanaged delegation.
Proposed table of contents
Chapters that map directly to the site authority pillars.
Ch. 1POV
The Network Perimeter Is Gone
Why agents break old assumptions about boundaries and trust.
Ch. 2Framework
The Seven Surfaces of AI Risk
The core framework and language for the enterprise.
Ch. 3NHI
Agent Identity and Delegated Authority
Why non-human identity becomes the hardest control problem.
Ch. 4Technical
Context Is an Attack Path
RAG, documents, memory, browser state, and indirect prompt injection.
Ch. 5MCP
Tool Use Changes Everything
MCP, SaaS actions, APIs, and the shift from text generation to action.
Ch. 6GRC
The Board-Ready AI Security Program
Governance, red teaming, metrics, and defensible reporting.
“The most dangerous AI systems in the enterprise will not look dangerous. They will look useful, connected, and under-governed.”
Pull quote placeholder for launch copy.
“The agent is not a user interface. It is a delegated actor. Secure it accordingly.”
Endorsement area / reviewer quote placeholder.
Executive offer
Bring The New Perimeter to your leadership team.
Use the book and framework as the basis for a board briefing, CISO workshop, or AI security posture assessment.