Agent Planning

Agent planning refers to the way an AI agent breaks a goal into a sequence of steps and decides what to do next, often by calling tools, querying systems, or taking actions with little or no human input at each step. For governance teams, it matters because the agent's chosen path, not just its final output, determines what data it touches, which systems it changes, and what risks it creates. Oversight therefore has to cover the planning and action steps themselves, such as logging decisions, setting limits on permitted actions, and defining when a person must approve a step.

1 item

ResearchGlobal2026-10-31

CrowdStrike Taxonomy Exposes Agent Planning Layer as an Untested Attack Surface

CrowdStrike published new research classifying prompt injection techniques that target the planning behavior of autonomous AI agents, not just their initial inputs or final outputs. The research shows that attackers can manipulate the intermediate decisions agents make while executing tasks. Enterprises that have tested only input and output layers may have a significant assurance gap in deployed agent systems.

CrowdStrike prompt injection agentic AI red teaming agent planning runtime monitoring adversarial testing