Use of customer data with Delegate and Cartographer

Data processing practices for Delegate and Cartographer, covering screen content, tool inputs, and conversation data processed to complete tasks and respond to users.

Purposes

We process the information used with our services for the following purposes:

• Provide our services

Provide our services

We process user inputs — including screen content, screenshots, documents, and tool output — provided at runtime to deliver our services, such as completing tasks and responding to the user with a recommended output.

We process telemetry data such as usage data to make sure our platforms operate smoothly and to optimize the user experience based on the feedback and usage patterns we observe. Telemetry carries operational signals only and never includes conversation content, tool output, or screenshots.

We use data analytics to measure and evaluate the performance of our services, as well as to improve the functionalities.

We process account information such as cloud ID, username, and password in accordance with Automation Cloud policies in order to provide you with access to the services and to secure your account.

We may process data that you send to us for support or as necessary for troubleshooting issues, in accordance with our Support terms.

We do not use your data to train the LLM models.

Data storage

By default, conversation history is stored in UiPath Automation Cloud. A local-only mode keeps that history on your machine instead.

Data is stored in accordance with our Data Residency page, available in the Automation Cloud admin guide.

Please review Delegate - Data security and compliance for additional information about conversation data storage and retention.

Integration with generative AI services

Delegate and Cartographer use the models made available through your organization's UiPath AI Trust Layer to process user inputs and generate responses — see Choosing the LLM model. These LLM model requests are governed by our AI Trust Layer.

Data is not used to train third-party models when using Delegate and Cartographer.

Safeguards

We employ the following safeguards in achieving the previously described purposes:

• Delegate and Cartographer functionality can be configured centrally through Automation Ops.
• An always-on redaction layer strips secrets and personal data from tool output before it reaches the model. See Delegate - Data security and compliance.
• The customer can delete conversation history directly through the Delegate and Cartographer interface, or by asking Support for a deletion request.