CVE-2026-76395

Remote Code Execution (RCE) through Deserialization of Untrusted Data in the Model Loading REST API in Splunk AI Toolkit

Published Aug 19, 2026 CVSS 3.1 · 8.8 High EPSS < 1% KEV · Not listed
Description
In Splunk AI Toolkit versions below 6.0.0, a user who holds the "power" Splunk role could execute arbitrary code on the Splunk server by loading a model file containing crafted sparse matrix data. The deserialization of untrusted data is possible because a model codec in Splunk AI Toolkit deserializes sparse matrix data without guarding against embedded pickle content. For more information see Troubleshoot the Splunk Machine Learning Toolkit (https://help.splunk.com/en/splunk-cloud-platform/apply-machine-learning/machine-learning-toolkit-user-guide/5.5.0/troubleshooting-mltk/troubleshoot-the-splunk-machine-learning-toolkit) in the Splunk documentation.

Remediation

Vendor Solution

Upgrade each affected Splunk app or add-on to the applicable fixed version listed in Product Status.


Vendor Workaround

Turn off or remove the Splunk AI Toolkit app. For more information see [Manage app and add-on objects](https://help.splunk.com/en/splunk-enterprise/administer/admin-manual/10.2/meet-splunk-apps/manage-app-and-add-on-objects) in the Splunk documentation. Note: turning off Splunk AI Toolkit turns off AI Toolkit Search Processing Language (SPL) commands and model operations. Splunk App for Data Science and Deep Learning and custom search commands that depend on AI Toolkit models or APIs might stop functioning. Unrelated Splunk custom search commands are not affected. For more information see [Troubleshoot the Splunk Machine Learning Toolkit](https://help.splunk.com/en/splunk-cloud-platform/apply-machine-learning/use-ai-toolkit/5.5.0/troubleshooting-mltk/troubleshoot-the-splunk-machine-learning-toolkit) and [DSDL install/version dependencies](https://help.splunk.com/en/splunk-enterprise/apply-machine-learning/splunk-app-for-data-science-and-deep-learning/5.1/install-and-configure-the-splunk-app-for-data-science-and-deep-learning/install-or-upgrade-the-splunk-app-for-data-science-and-deep-learning) in the Splunk documentation.

Tracking

Sign in to view the affected projects.

Advisories

No advisories available
History
Type Values Removed Values Added
Metrics ssvc

{'options': {'Automatable': 'no', 'Exploitation': 'none', 'Technical Impact': 'total'}, 'version': '2.0.3'}

Type Values Removed Values Added
First Time appeared Splunk ai Toolkit
CPEs cpe:2.3:a:splunk:ai_toolkit:*:*:*:*:*:*:*:*
Vendors & Products Splunk ai Toolkit
Type Values Removed Values Added
First Time appeared Splunk
Splunk splunk Ai Toolkit
Vendors & Products Splunk
Splunk splunk Ai Toolkit
Type Values Removed Values Added
Description In Splunk AI Toolkit versions below 6.0.0, a user who holds the "power" Splunk role could execute arbitrary code on the Splunk server by loading a model file containing crafted sparse matrix data. The deserialization of untrusted data is possible because a model codec in Splunk AI Toolkit deserializes sparse matrix data without guarding against embedded pickle content. For more information see Troubleshoot the Splunk Machine Learning Toolkit (https://help.splunk.com/en/splunk-cloud-platform/apply-machine-learning/machine-learning-toolkit-user-guide/5.5.0/troubleshooting-mltk/troubleshoot-the-splunk-machine-learning-toolkit) in the Splunk documentation.
Title Remote Code Execution (RCE) through Deserialization of Untrusted Data in the Model Loading REST API in Splunk AI Toolkit
Weaknesses CWE-502
References
Metrics cvssV3_1

{'score': 8.8, 'vector': 'CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H'}

Weaknesses

Sources

Status
PUBLISHED
Assigner
cisco
Published
Updated
Aug 26, 2026, 7:51 PM
Reserved
Aug 19, 2026, 12:02 PM
Status
Analyzed
Published
Aug 19, 2026, 10:17 PM
Modified
Aug 26, 2026, 8:18 PM