# Access Policies

## Create & Deploy Policy

* To create an AI Agent access policy, navigate to the desired AI Agent and go to the **Access Policies** tab
* Click **Create Policy** and provide a name and description, then click **Next.**
* Next, select the desired **Resource Source** for the policy

{% hint style="warning" %}
Currently, only Data Sources & Data Products are supported discoverable resource sources.
{% endhint %}

* After selecting the desired **Resources**, click **Next**. 
* Policy conditions can be made with respect to **Resource Attributes** and **User Attributes/Entitlements**
* After saving and deploying the policy, the policy will be evaluated and decision returned dynamically to the AI Agent when it requests access to the resource via MCP. See [mcp-server](https://docs.trustlogix.io/trust-ai-private-preview/mcp-server "mention") for more details.