2026-07-15

Agent Autonomy Isn't the Risk, Loss of User Authorization Is

Insurance carriers just walked away from covering AI incidents. Not because AI is too risky to insure, but because most agentic systems can't answer whose authority actions are being performed.

Read post

2026-07-06

Where Agentic Architecture Has To Go

The security requirements of agentic systems are pointing us toward an architectural outcome nobody is talking about yet.

Read post

2026-03-20

The Invisible Break

Autonomous agents break the link between authorization and execution, enterprises are deploying them anyway.

Read post