# CVE-2026-11876: zenml-io ZenML missing authorization in stack deployment endpoint

Source: https://junglewise.ai/threats/cve-2026-11876-zenml-io-zenml-missing-authorization-in-stack-deployment-endpoint  
Publisher: Junglewise Threat Intelligence, https://junglewise.ai/threats  
Published: 2026-07-21. Updated: 2026-07-21  
License: CC BY 4.0 with a link back

CVE-2026-11876 · Severity: medium · CVSS 5 · Published 2026-07-21

ZenML is an open-source framework used to manage machine learning workflows and infrastructure. A security flaw in its management interface allows any logged-in user to view technical details about infrastructure setups (stacks) belonging to other users or organizations. This could allow an unauthorized person to map out a company's internal cloud environment, potentially leading to further targeted attacks or data exposure.

Technologies: [ZenML](https://junglewise.ai/threats/technologies/zenml). Vendors: [ZenML](https://junglewise.ai/threats/vendors/zenml).

## Quick answers

**Is CVE-2026-11876 exploited in the wild?** No. Junglewise Threat Intelligence has no report of CVE-2026-11876 being exploited in the wild.

**How severe is CVE-2026-11876?** Medium severity, with a CVSS score of 5, published 21 July 2026.

**Which products and versions are affected?** ZenML 0.94.2 and all versions prior to 0.95.0.

**Is there a fix for CVE-2026-11876?** A patch was published on 2026-07-21: Fix committed to GitHub repository Junglewise finds every place you run it, drafts the upgrade and applies it once you approve.

**How is CVE-2026-11876 attacked?** It is attacked over the network, in a software library.

**Am I affected by CVE-2026-11876?** Check whether you run ZenML at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

## Technical details

A missing authorization vulnerability (CWE-862) exists in the `GET /api/v1/stack-deployment/stack` endpoint within ZenML. The `get_deployed_stack` function fails to implement Role-Based Access Control (RBAC) checks and utilizes a server-side `Client()` that interacts directly with `SqlZenStore`, bypassing the standard RBAC enforcement layer. An authenticated attacker can exploit this to enumerate all deployed stacks, exposing sensitive metadata including infrastructure topology, service connector details, and owner user IDs. This is particularly impactful in multi-tenant ZenML Pro or Cloud environments where cross-tenant reconnaissance is possible. The issue is addressed in version 0.95.0.

## Affected products

- zenml-io ZenML 0.94.2 and all versions prior to 0.95.0

## Timeline

- 2026-07-21: advisory: NVD publication date
- 2026-07-21: disclosed: Vulnerability reported via huntr.dev
- 2026-07-21: patched: Fix committed to GitHub repository

## References

- [https://github.com/zenml-io/zenml/commit/9ad2c65f24ded22a5a98d289d63a7f629b434b61](https://github.com/zenml-io/zenml/commit/9ad2c65f24ded22a5a98d289d63a7f629b434b61)
- [https://huntr.com/bounties/3b434e02-0ee9-4b5e-a44b-0988b6f074c7](https://huntr.com/bounties/3b434e02-0ee9-4b5e-a44b-0988b6f074c7)

## Related threats

- [ZenML rate limit bypass in login and password-change endpoints](https://junglewise.ai/threats/cve-2026-11922-zenml-rate-limit-bypass-in-login-and-password-change-endpoints)
- [ZenML reflected cross-site scripting in survey redirect](https://junglewise.ai/threats/cve-2024-5062-zenml-reflected-cross-site-scripting-in-survey-redirect)