Arbitrary Code Execution Risk in Giskard Framework for AI Models
CVE-2026-40320

5.4MEDIUM

Key Information:

Vendor

Giskard-ai

Status
Giskard-oss
Vendor
CVE Published:
17 April 2026

What is CVE-2026-40320?

The Giskard framework, an open-source tool for AI model testing, is susceptible to a serious vulnerability in the ConformityCheck class. In versions prior to 1.0.2b1, it improperly handles the rendering of the rule parameter by utilizing Jinja2's Template constructor. This oversight allows untrusted sources to execute crafted rule strings, potentially leading to arbitrary code execution. Exploitation of this vulnerability necessitates write access to the check definition followed by the execution of the test suite. This issue has been addressed in the recent update to version 1.0.2b1.

Affected Version(s)

giskard-oss < 1.0.2b1

References

https://github.com/Giskard-AI/giskard-oss/security/adviso...
x_refsource_CONFIRM
https://github.com/Giskard-AI/giskard-oss/releases/tag/gi...
x_refsource_MISC

CVSS V4

Score:
5.4
Severity:
MEDIUM
Confidentiality:
High
Integrity:
High
Availability:
High
Attack Vector:
Local
Attack Complexity:
Low
Attack Required:
Physical
Privileges Required:
Undefined
User Interaction:
Unknown

Timeline

  • Vulnerability published

  • Vulnerability Reserved

.