Unsafe Delegation
Testing how authority and identity propagate when one agent delegates a sub-task to another agent or automated process.
- Category
- Agentic AI Security
- Difficulty
- Advanced
- Last Updated
- 20 September 2026
Lab detail
Objective
Multi-agent and agent-to-automation designs delegate sub-tasks downstream, and that delegation has to carry (or deliberately not carry) the original identity, authority and constraints. This lab will build a small reference two-agent chain and test what happens when a delegated sub-task is manipulated, or when the receiving agent trusts the delegating one more than it should.
Planned Scope
- Build a reference two-agent (or agent-to-automation) delegation chain with a defined task.
- Test whether the delegated identity/authority is scoped correctly at each hop.
- Test manipulation injected at the point of delegation, not just at the original entry point.
- Document delegation patterns (explicit scoping, re-authorization, audit trail) that hold up.
Framework Mapping
Referenced frameworks
Exact control-ID mappings are to be validated as this lab progresses — see the full Frameworks reference.
Responsible Use / Safety Notes
How this lab is run
This lab uses synthetic data and controlled environments. It is intended to demonstrate defensive AI-security principles and does not contain customer data, production credentials or intentionally weaponized exploit infrastructure.
A decision is waiting for sign-off. Get an independent view first.
Tell me what is being decided and what is at stake. I will first say honestly whether my experience fits — then we talk about scope, not before.
Or connect on LinkedIn first.