For most of my career in cybersecurity, one assumption has always been there: behind every important action, access request, transaction, or decision, there was ultimately a person.
That assumption is beginning to disappear.
AI is moving beyond answering questions.
AI agents can pursue objectives, use tools, access systems, communicate with other agents, make decisions, and act on behalf of an organization.
And this changes much more than technology.
It changes security, governance, accountability — and eventually the very structure of the enterprise.

That is why I wrote:
📘 The Last Human Workforce
Beyond Agentic AI: The Rise of Autonomous Organizations
The book explores what may happen as we move from:
AI → AI Agents → Agent Ecosystems → Autonomous Organizations

And it introduces several concepts that I believe will become increasingly important:
The Rented Mind: when thousands of agents depend on intelligence the organization does not own.
The Borrowed Hands:when agents act through APIs, tools and platforms the organization does not fully control.
The Vanishing Human: when a human defines an objective, but thousands of autonomous decisions follow.
And one principle sits behind all of them:
The Principle of Least Agency.
We spent decades implementing least privilege.
In the age of Agentic AI, that may no longer be enough.
We will also need to decide how much autonomy a machine should be allowed to have.
The question is no longer simply:
What can AI do?
The more important question may be:
How far are we prepared to let it act?
📘 The Last Human Workforce is now available on Amazon.