GTM.ai
Docs menu

How to acknowledge the AI agent disclosure before running an agent team

Learn how to acknowledge the AI agent disclosure the first time you run or activate an agent team.

Srivatsa MarthiSrivatsa MarthiPrincipal Product Manager

Learn how to acknowledge the AI agent disclosure the first time you run or activate an agent team.

Overview

Agent teams use AI agents that can take actions on your behalf. When your organization has this turned on, the first time you run or activate an agent team, GTM One shows a dialog titled "Enabling AI Agents" and asks you to acknowledge it. The run or activation does not start until you do. Once you have acknowledged it, your answer is saved to your account, and the dialog does not interrupt your runs again.

The dialog can appear when you run an agent team, activate one, retry runs that were rejected, or start one from a document, an Audience, or another place in GTM One. It is the same dialog everywhere.

This is available when your organization has it turned on. If you never see the dialog, the feature is not on for your organization.

Acknowledge the Disclosure and Continue

  1. Run or activate an agent team (e.g., open Agent Teams in the left navigation panel and start a run, or activate a team from the list).
  2. In the "Enabling AI Agents" dialog, read the text. It explains that the agents act autonomously within the permissions granted, that their outputs and actions may be inaccurate or unexpected, and that some actions may not be reversible.
  3. To read the full details first, click Learn more about AI agent risks, read the seven sections of "Before You Enable Agent Teams", then click Back.
  4. Select the checkbox I understand and want to enable these AI agents.
  5. Click Continue.

The run or activation you started then goes ahead. Continue stays turned off until the checkbox is selected.

To leave without running anything, click Cancel. Nothing runs, and the dialog appears again the next time you run or activate an agent team. To see the agent teams list, click View Agent Teams in the dialog. The list opens in a new browser tab, and the dialog stays open in the first tab.

What the Details Cover

The "Before You Enable Agent Teams" view has seven sections:

  • You are in control: you can review or revoke the permissions and disable the feature at any time from the Agent Teams list page, and disabling it does not undo actions already taken.
  • AI-driven behavior: the agents interpret instructions and context, so their behavior may not always match what you intended.
  • Review permissions carefully: the agent uses the full scope of access you provide, so grant only the access it needs.
  • Some actions may be irreversible: for example sending communications or overwriting data. Consider requiring human review or approval before high-impact actions.
  • Accuracy and reliability: outputs may be inaccurate, so do not rely on them alone for high-stakes decisions.
  • Data use and access: the agent may use data you have enabled, such as contact records, calendar information, email content, and CRM data. Review new uses of that data with your legal counsel.
  • Human oversight: review the agent's activity log and outputs regularly. You remain responsible for actions taken using your account.

If you see an error

  • "We couldn't save your acknowledgment. Please try again." Your acknowledgment was not saved, so the run did not start. Click Continue again, and contact your ZoomInfo admin if it keeps failing.

Resources