Healing Agent user guide

Last updated May 9, 2025

Enabling Healing Agent using the AI Trust Layer policy

Healing Agent operates according to the AI Trust Layer configuration, which allows you to align its decisions and actions to the defined level of trust. You can therefore balance automation flexibility against precision, reduce false positive risk, and maintain confidence in the self-healing operations.

For example, if your processes use confidential data, you can use AI Trust Layer to opt out of AI-based automation recovery. In this case, Healing Agent only uses non-AI-based, heuristic recovery methods for recommendations and automation healing.

To perform AI-based recovery strategies, Healing Agent issues calls to third-party AI models. Healing Agent uses AI calls for text matching.

Enabling or disabling Healing Agent

To control access to Healing Agent, follow these steps:

Navigate to Automation Ops™ in your Automation Cloud™ organization.
Create or edit a product policy for AI Trust Layer.
For details, refer to Creating a governance policy.
On the Product Toggles tab, activate the Enable Healing Agent option.
Save the policy.
Deploy the policy to any tenant where you want to build or run agents.