Notifications
Clear all
Topic starter
18/08/2026 2:13 pm
TL;DR: Models can score well on benchmarks and still fail as agents, because production reliability depends on the harness, not the model, according to Fiddler. The article argues that system instructions, context assembly, guardrails, sandboxes, and human escalation points determine whether agents fail safely, especially when tool calls, permissions, and long-tail cases go wrong.
NHIMG editorial — based on content published by Fiddler: Building Reliable Agents Takes More Than Better Models
By the numbers:
- Only 44% of organisations have implemented policies to govern AI agents, even though 92% say governing them is critical to enterprise security.
- 80% of organisations report their AI agents have already performed actions beyond their intended scope, including accessing unauthorised systems, sharing sensitive data, or revealing credentials.
- When AWS credentials are exposed publicly, attackers attempt access within an average of 17 minutes, and as quickly as 9 minutes in some cases.
Questions worth separating out
Q: What breaks when AI agents are given broad standing access?
A: Broad standing access breaks governance because the agent can move from one task to another without a fresh authorization check.
Q: Why do AI agents need identity and access governance if the model is already strong?
A: Because model strength does not control who or what the agent can touch in production.
Q: How do security teams know whether an AI agent is operating safely?
A: Security teams know an AI agent is operating safely when its permissions, invoked tools, and accessed data remain consistent with the approved use case over time.
Practitioner guidance
- Instrument the agent before enabling autonomy Add logging, traceability, and control points before connecting the agent to live tools or business data.
- Redesign tool interfaces for minimal context Expose only the fields the agent needs for the next step, and remove broad passthrough responses that flood the context window.
- Bind permissions to task scope and runtime approval Use least privilege for agent credentials, and require escalation before actions that touch sensitive systems, credentials, or high-impact outputs.
What's in the full article
Fiddler's full blog covers the deployment mechanics this post intentionally leaves at a higher level:
- A deeper explanation of how harness design changes reliability across system instructions, memory, execution loops, and sandboxes.
- Examples of where agent failures cluster in production, including tool calls, permissions, output formatting, and long-tail edge cases.
- A closer look at how teams should instrument agents before enabling autonomy, including the observability needed for review and escalation.
- The practical discussion from the AI Explained session with Juhi Parekh on how frontier-model teams think about deployment failure modes.
👉 Read Fiddler's analysis of why reliable AI agents depend on the harness, not just the model →
AI agent reliability: what changes when the harness does the work?
Explore further
18/08/2026 3:05 pm
Agent reliability is an identity governance problem disguised as a model engineering problem. Once an AI system can act across tools and permissions, the central question becomes who or what is authorised to do which action, on which data, and under what supervision. That puts IAM, PAM, and NHI controls on the critical path for agentic AI deployments. The practitioner conclusion is simple: treat the agent as a governed identity surface, not a chat interface.
A question worth separating out:
Q: Should organisations treat AI pentesting agents like non-human identities?
A: Yes. If an agent can authenticate, browse, test, and interact with systems, it has an authority boundary that should be governed like any other non-human identity. That means ownership, least privilege, explicit scope, isolation, and revocation controls. The difference is that these controls must be runtime-enforced, not assumed from policy.
👉 Read our full editorial: Building reliable AI agents requires harness-level controls
