Data processing

To power AI Autofill inside a playbook step, all of the data available at that point in a run may be transmitted to and processed by the ChatGPT API by OpenAI. This includes, for example:

• All data from the trigger
  • e.g. if you use a Gmail trigger: the subject, body, recipients, etc.
  • or if you use a Calendly trigger: the names & email addresses of all attendees, the date & time of the event, etc.
• All output data from previous automations
  • e.g. if you added a row to a Sheet: all of the column values
  • or if you added a new issue in Jira: the issue title, description, assignee, epic, etc.
• All output data looked up from another app
  • e.g. if you looked up a deal from Hubspot: the deal name, stage, estimated value, probability, close date, etc.
  • or if you looked up a row in Airtable: all of the columns in the row

As per its API policy, OpenAI does not train on the data or retain the data transmitted from Relay to OpenAI.

There is currently no way to exclude specific data from being transmitted to OpenAI if you use the AI Autofill feature in a playbook. If your playbook handles any sensitive data that you do not want to be transmitted to or processed by OpenAI, we advise you to not use the feature, or to use it from a different playbook that does not handle any sensitive data.

If you do not use the feature in your playbook, the ChatGPT API will not be called with your data.