Notifications
Clear all
Topic starter
02/09/2026 1:29 pm
TL;DR: Agent risk is shifting from the model to the harness, the orchestration layer that turns tokens into tool calls, file writes, and API actions, according to Island. The practical implication is that teams must govern visibility and enforcement where agent authority is actually exercised, not only at the model layer.
NHIMG editorial — based on content published by Island: The Harness Dilemma, why model guardrails aren't enough for agent security
Questions worth separating out
Q: What breaks when AI agents are not governed at runtime?
A: Without runtime governance, an agent can shift behaviour after provisioning and still execute actions that were never reviewed in context.
Q: Why do agent harnesses change the risk profile of the same model?
A: The harness determines how context is assembled, which tools are available, and whether decisions are rechecked before action completes.
Q: How do security teams know if an AI agent is operating outside its approved role?
A: Teams should compare actual workflow behaviour against the approved use case.
Practitioner guidance
- Map the harness as a governed control surface List every production agent, the harness that wraps it, the tools it can invoke, and the identities or secrets it can reach.
- Separate model approval from runtime authorisation Approve models for quality and safety, but enforce tool use, file access, and API actions with policy outside the reasoning loop.
- Correlate prompts, tool calls, and identity Build visibility that reconstructs the full path from prompt to response to tool invocation to file touch and credential use.
What's in the full article
Island's full blog covers the operational detail this post intentionally leaves for the source:
- Specific examples of harness failures across automation repositories and red-team tests
- The full visibility model for prompt, tool call, file access, and identity correlation
- Detailed runtime enforcement patterns across browser, desktop, network, and MCP gateway layers
- The practical distinctions between install-time scanning and live policy enforcement
👉 Read Island's analysis of harness security and agent governance →
Harness security for AI agents: are your controls keeping up?
Explore further
02/09/2026 1:30 pm
Harness security is becoming the real agent governance boundary. Model selection matters, but the harness decides whether output stays advisory or becomes action. That means the enterprise control plane for agentic AI is moving toward orchestration, tool mediation, and policy enforcement outside the model loop. For identity teams, the question is no longer only what the agent knows, but what identity and privilege it can exercise at runtime.
A question worth separating out:
Q: When should organisations treat an AI system as a non-human identity?
A: Treat an AI system as an NHI when it can authenticate, request tools, or perform actions without direct human supervision. At that point it needs inventory, lifecycle, least privilege, monitoring, and revocation controls just like other machine identities.
👉 Read our full editorial: Harness security is now the control plane for AI agents
