Label Resources Using AI Labeling
AI Labeling simplifies resource labeling by automatically discovering resources across cloud and data center environments and generating label recommendations. After you review and approve a recommendation, the feature assigns the appropriate labels without requiring manual analysis of each workload or resource.
AI Labeling streamlines the process by generating recommendations automatically and assigning labels after your approval. This reduces manual effort while helping ensure labels are applied consistently to the selected resources.
How recommendations are generated
Illumio AI Labeling learns from your environment to recommend workload labels. Recommendations are based on your organization's conventions and on an analysis of the resource metadata, cloud tags, hostnames, running processes, and existing labeling patterns in your environment.
AI Labeling considerations
AI Labeling supports standard Illumio label types (Role, Application, Environment, and Location), with only one label of each type allowed per resource; Role label recommendations apply only to AWS instances.
Recommendations are generated approximately once per day, and resources remain eligible even if tag-to-label mappings exist but have not been approved or applied.
Approving a recommended label automatically applies any existing security policies that reference the same label value.
Depending on the volume, label assignment may take some time and will continue in the background and completes with a notification.
AI Labeling workflow
You can inspect recommended labels at both a summary and resource level, then approve or ignore recommendations in bulk or individually to let AI Labeling assign labels across your environment.
STEP 1: View and filter recommendations
Go to Label Management > Labeling Method > AI Labeling.
The AI Labeling page automatically displays label values discovered across your cloud and data center resources. By default, the results are shown in the Recommended tab and filtered by the Application label type.
Click View Details in the label recommendation card or list view (see icons) to see the cloud and data center resources that are candidates for the recommended label.
Filter the recommendations for a more granular view. Filtering options include:
Tiles or list. Click an icon to switch between tile and list view.
Group by one of the four Illumio label types/dimensions (Role, Application, Environment, Location).
Recommended Labels. Filter by one or more of the AI-recommended values for the label type selected in the Group by filter.
Time. Specify a time range for when recommendations were generated.
Sort by resource count, label value, or time.
STEP 2: View label recommendation evidence
Go to the Details page by clicking View Details in the label recommendation card or list view.
Click the flask icon
in the Evidence column to open the slide-out.
STEP 3: Approve or ignore recommendations
Approve or ignore recommendations. There are two ways:
In an individual resource's slide-out panel: This is another way to approve or ignore a single recommendation. From the Details page, click the flask icon
adjacent to the listed resource to open the slide-out. Click Approve or Ignore. Follow the onscreen prompts to complete the action.On a recommendation's Details page: This method is convenient for approving or ignoring recommendations in bulk. Select one or more resources and then click Approve or Ignore. Follow the onscreen prompts to complete the action.