September 2026

September 2026 release notes for Automation Ops in Automation Cloud and Test Cloud.

September 16, 2026

Manage governance policies and CI/CD pipelines with a coding agent — Public Preview

You can now author governance policies and manage CI/CD pipelines in Automation Ops entirely from a coding agent (Claude Code, Codex, Cursor, and others) and terminal.

Two skills cover this:

• uipath-governance — authors and deploys product and access policies, and checks compliance posture against a standard such as ISO 42001 or ISO 27001, then diagnoses why a policy isn't taking effect.
• uipath-aops — discovers Studio Admin source control state, and creates, runs, and monitors CI/CD pipelines that build a source-controlled Studio project and update a published Orchestrator process.

Both skills are available to install via uip skills install.

For details, refer to the following pages:

• Governance policies
• CI/CD pipelines
• UiPath for Coding Agents — Install and set up
• UiPath for Coding Agents — September 2026 release notes

September 9, 2026

Agent design policies in Studio Web are now generally available

The agent design policies in the Automation Ops governance policy for Studio Web are now generally available. The policies let you enforce the following agent configuration requirements before an agent is published:

• Minimum agent score
• Maximum temperature range
• Maximum tokens per agent response
• Human-in-the-loop escalation
• Maximum iterations per agent execution

For details, refer to Settings for Studio Web policies.

September 7, 2026

Correction, September 17, 2026: This update was released on September 7, 2026, but was not included in the release notes at the time.

New AI Trust Layer product toggles for Context Grounding and the LLM as Judge guardrail

The AI Trust Layer policy now includes two product toggles: Enable Context Grounding and Enable LLM as Judge guardrail. The toggles control whether Context Grounding and the LLM as Judge guardrail can make calls to AI models through the AI Trust Layer. Both are set to Yes by default, so existing policies keep their current behavior and Context Grounding usage continues uninterrupted.

For details, refer to Settings for AI Trust Layer policies.