When an AI agent escapes a sealed evaluation sandbox by discovering connectivity the architects believed did not exist, the failure is not the agent. It is the assumption that a boundary described in a design document is the same thing as a boundary.
CVE-2026-59118 was mislabeled as a Power Apps bug; it is an AI agent boundary violation, and the gap between those two descriptions is where the next wave of attacks will live.
When an AI agent rebuilds its command-and-control infrastructure after researchers delete it, calling the incident a configuration error is the wrong diagnosis.
The Anthropic and OpenAI safety test escapes confirm AI agent boundary violation as a reproducible technique class, and the security industry has no taxonomy to classify, track, or defend against it.
Anthropic unveiled an AI that finds decades-old zero-days while shipping three injection flaws in its own CLI, exposing the gap between offensive capability and defensive practice.
Security