Skip to content

September 8, 2026 - On-premises - v10.102.0.7

This release includes updates in the following areas for F5 AI Security:

Moderator versions

AppVersion
Moderatorv10.102.0.7
Workflowsv1.138.1

Helm chart versions

Chart NameVersion
f5-ai-security-operator-helm1.8.0
cai-moderator1.13.0
cai-prefect1.8.0

Breaking changes

Inference model pods must run on nodes with NVIDIA driver version 580 or later. For more information, see the latest version of the NVIDIA Driver Installation Guide.

On-premises

AI Security on-premises deployments now include all of the features described in the following SaaS releases:

With this release, you can now receive notification in product when a license is approaching its expiry date. You can view license details and expiry dates from the Account Settings page, and can self-serve enter a new license key into the product. You can also view your license status and expiration date with the GET /backend/v1/license API endpoint. This endpoint is unauthenticated and available for on-premises deployments. For more information, see AI Security license management.

Resolved issues

  • For legacy projects created before May 20th: when a user is assigned as a project admin, they got Organisation-{orgID}-manage-api-tokens. In other words, such project admins saw all cards on their project list page.
    • Project data remained secure. Selecting any unauthorized card yields a 404.
  • Any user assigned this project admin role were also able to create API tokens at the org level, which gave them full Organization admin access.

Doc updates

We have published the following updates to our product documentation since the last on-premises release on July 21, 2026:

Updated at: