# CVE-2026-45758: Guardrails AI embedded malicious code in PyPI package

Source: https://junglewise.ai/threats/cve-2026-45758-guardrails-ai-embedded-malicious-code-in-pypi-package  
Publisher: Junglewise Threat Intelligence, https://junglewise.ai/threats  
Published: 2026-06-05. Updated: 2026-06-05  
License: CC BY 4.0 with a link back

CVE-2026-45758 · Severity: critical · CVSS 9.6 · Published 2026-06-05

Guardrails AI, a framework for building AI applications, was the victim of a supply chain attack where a malicious version of the software was uploaded to the official Python package registry (PyPI). Any user who installed version 0.10.1 on May 11, 2026, may have had their system compromised, potentially leading to the theft of sensitive credentials like cloud provider keys and API tokens. Organizations should immediately downgrade to version 0.10.0 or upgrade to 0.10.2 and rotate all secrets accessible from affected development environments.

Technologies: [guardrails-ai (PyPI)](https://junglewise.ai/threats/technologies/pypi-guardrails-ai). Vendors: [PyPI](https://junglewise.ai/threats/vendors/pypi).

## Quick answers

**Is CVE-2026-45758 exploited in the wild?** No. Junglewise Threat Intelligence has no report of CVE-2026-45758 being exploited in the wild.

**How severe is CVE-2026-45758?** Critical severity, with a CVSS score of 9.6, published 5 June 2026.

**Which products and versions are affected?** Guardrails AI 0.10.1; guardrails-ai (PyPI) ECOSYSTEM:introduced=0,last_affected=0.10.1.

**Is there a fix for CVE-2026-45758?** No fixed version is recorded yet. Check the vendor's advisory in the references below. If you run it, Junglewise drafts a mitigation that fits your setup and applies it once you approve, with senior security engineers behind it.

**How is CVE-2026-45758 attacked?** It is attacked over the network, in a software library.

**Am I affected by CVE-2026-45758?** Check whether you run Guardrails AI or guardrails-ai (PyPI) 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 supply chain compromise resulted in the publication of guardrails-ai version 0.10.1 containing embedded malicious code (CWE-506). The malicious logic was injected into 'guardrails/__init__.py' and is designed to automatically fetch and execute a remote Python artifact on Linux systems when the package is imported. The attack is attributed to the threat actor 'TeamPCP'. While the package was quarantined by PyPI within two hours, users who successfully installed it are at risk of full system compromise and credential exfiltration. Remediation requires downgrading to 0.10.0 or upgrading to 0.10.2, followed by a comprehensive audit of environment credentials.

## Affected products

- Guardrails AI guardrails-ai 0.10.1

## Identifiers

- CVE-2026-45758
- GHSA-xmpw-2vmm-p4p6
- PYSEC-2026-206

## Timeline

- 2026-05-11: disclosed: Malicious version 0.10.1 published to PyPI.
- 2026-05-11: other: PyPI quarantined the malicious repository.
- 2026-05-12: advisory: Security advisory and tracking issue published by Guardrails AI and Socket.
- 2026-06-05: other: CVE-2026-45758 published to NVD.

## References

- [https://github.com/guardrails-ai/guardrails/blob/main/SECURITY_ADVISORY.md](https://github.com/guardrails-ai/guardrails/blob/main/SECURITY_ADVISORY.md)
- [https://github.com/guardrails-ai/guardrails/issues/1473](https://github.com/guardrails-ai/guardrails/issues/1473)
- [https://github.com/guardrails-ai/guardrails/security/advisories/GHSA-xmpw-2vmm-p4p6](https://github.com/guardrails-ai/guardrails/security/advisories/GHSA-xmpw-2vmm-p4p6)

## Related threats

- [PYSEC-2026-1432 - Guardrails has an arbitrary code execution vulnerability](https://junglewise.ai/threats/cve-2024-45858-guardrails-has-an-arbitrary-code-execution-vulnerability)
- [PYSEC-2026-1431 - Guardrails AI vulnerable to Improper Restriction of XML External Entity Reference](https://junglewise.ai/threats/cve-2024-6961-guardrails-ai-vulnerable-to-improper-restriction-of-xml-external)
- [Guardrails AI code injection in Hub package installation](https://junglewise.ai/threats/cve-2026-31233-guardrails-ai-code-injection-in-hub-package-installation)