Topic starter
07/03/2026 9:53 pm
TL;DR: AI coding agents are moving into the SDLC with repository, pipeline, and production access, which turns security review tools into privileged identities that can modify builds, leak code, or trigger deployments if compromised, according to Token Security. Existing IAM models were built for human-paced provision, authenticate, deprovision cycles, and that assumption breaks when an agent authenticates continuously and acts task-by-task.
Editorial analysis by NHI Mgmt Group, based on content published by Token Security: “Claude Code Security Is A Leap Forward for Application Security. But Who Governs It?”.
By the numbers:
- Non-human identities already outnumber human users 82 to 1.
Key questions
Q: What breaks when AI systems inherit broad repository access?
A: Broad inherited access lets AI systems reach data that was never intended for machine-scale retrieval, including stale, duplicated, or sensitive content.
Q: Why do AI coding agents create governance risk even when they improve productivity?
A: They create risk because faster output does not guarantee safer output.
Q: What do teams get wrong about AI agent access reviews?
A: Teams often assume an access review can certify an agent the same way they certify a human or a service account.
Practitioner guidance
- Classify AI coding agents as privileged identities Inventory every code-review, SAST, and repository-connected agent as a non-human identity with named ownership, explicit scope, and a revocation path.
- Replace standing permissions with task-bound access Scope repository, pipeline, and release permissions to the specific review task and terminate them when that task ends, rather than inheriting broad team roles.
- Separate detection from authorization Use code analysis and prompt filtering as detection layers, but do not treat them as substitutes for access governance or privilege reduction.
Bottom line: AI coding agents change the SDLC risk model because they can act as privileged identities inside repositories, pipelines, and release paths.
Explore further
View Full Forum → | NHI Foundation Course → | Our Services → | Read the full analysis →
This topic was modified 6 days ago by NHI Mgmt Group
06/10/2026 8:40 pm
AI coding agents have crossed the line from tooling to identity. Once a review system can authenticate into repositories, pipelines, and production context, it behaves like a privileged non-human identity, not a passive application. That means the governance problem is no longer only software quality or model safety. It is access scope, ownership, and lifecycle control for an actor that can affect delivery outcomes directly.
A few things that frame the scale:
- Claude Code-assisted commits leaked secrets at a rate of 3.2%, more than double the human-only baseline of 1.5%, with peaks reaching 31 secrets per 1,000 commits in August 2025, according to the State of Secrets Sprawl 2026.
- Companies are dedicating an average of 32.4% of their security budgets to secrets management and code security, with US organisations leading at 40.8%, according to the State of Secrets in AppSec.
A question worth separating out:
Q: How should organisations govern AI coding agents in the SDLC?
A: Treat them as privileged non-human identities with named owners, narrow task scope, and explicit offboarding. Governance should align access to the exact development activity, not to a broad team role or permanent environment trust. That is the difference between using the agent safely and leaving it as an unowned execution path.
👉 Read our full editorial: Claude Code Security and the governance gap for AI coding agents