Working with the AI Workforce

The AI Workforce is a set of virtual employees that automatically gather intelligence about your environment and present it to you in a briefing. This briefing is compiled after each discovery cycle.

Permissions

• To view the AI Workforce's runs, your role needs the AI Agents permission View AI Workforce Runs, and you must have a global data scope. The predefined administrator roles have permissions; custom roles don't until an administrator enables them in the role editor. Users assigned to a data scope don't have access to these runs, regardless of their role.
• An administrator can stop or resume runs from System Settings.

What the AI Workforce Does

The AI Workforce is a built-in account that owns one scheduled task per Virtual Employee check, each with the cadence After every discovery cycle. The tasks are active out of the box, so the first briefings arrive after the first completed cycle. The Teacher and Onboarding guides are conversations and have no tasks.

• Nobody can sign in as the AI Workforce account, and it doesn't appear under Service Accounts.
• The account follows each agent's settings: an agent set to Paused skips its runs, and Enabled Read-Only suppresses its writes.
• The account's write actions are Off unless an administrator sets Shadow or Live for the agent. Even with the agent set to Live, the AI Workforce runs only a preview and never applies it. Your own runs of that agent keep the Live setting.

Viewing the AI Workforce's Runs

If your role has the required permissions, the AI Workforce's runs and schedules appear among your own throughout the AI Workforce page, labeled AI Workforce, and count in your run totals, tiles, and headline metrics:

• Virtual Workers tab: a job's Last run cell shows the AI Workforce badge when the newest run is the account's; in the drawer's schedule table, the account's task shows "AI Workforce" in the Created by column; a note under the roster says the view includes the AI Workforce's runs and schedules.
• Activity Log: each of the account's runs carries an AI Workforce badge in the Owner column and an Open conversation button.
• Notification Center: finished runs appear as entries with the badge and count toward Unread until you read them.
• Agent profile: the Last run, Latest briefing, and Schedules entries carry the AI Workforce badge, and the run counter reads "N runs recorded by you and the AI Workforce."

The AI Workforce's runs never appear in the Other Users' Runs tab.

[Screen capture: Control Panel roster with an AI Workforce-labeled last run and the roster note]

Continuing an AI Workforce Conversation

1. To open a previous conversation, click Open conversation on its Agent Log card, or Open report on its Notification Center entry. The conversation opens read-only: a banner explains why, and the message box is disabled.

   [Screen capture: Read-only banner with the Clone conversation control]

2. To continue the conversation as yourself, click Clone conversation. Axonius copies the whole conversation into your chat history with the title "(title) (copy)" and opens the copy, where follow-up questions work as in any chat.

Note: You can't clone a run that is still in progress. The banner shows the platform's message; try again when the run finishes.

Granting the View AI Workforce Runs Permission

1. Open the role in the role editor.
2. Under AI Agents, enable View AI Workforce Runs.
3. Click Save.

Stopping AI Workforce Runs

1. Go to System Settings > AI settings.
2. In the AI Workforce section, disable Run the AI Workforce after every discovery cycle. The option is on by default.
3. Click Save.

[Screen capture: AI settings page with the AI Workforce section]

When Run the AI Workforce after every discovery cycle is disabled, the discovery-cycle trigger skips every AI Workforce task and stops runs already in progress. Axonius retains per-agent settings, and enabling the option again resumes runs after the next discovery cycle. Axonius records the change in the Activity Logs page, with the user and the time.