Privacy and Security

Privacy and Security Overview

We champion user privacy and data security, ensuring that your interactions with Cetegra AI Chat are safeguarded with robust measures for private and secure AI engagements.

Data Privacy Principles

  • Ownership: All data, whether prompts, completions, or training data, belongs to you.
  • No Sharing: Your data is inaccessible to other customers, OpenAI, Microsoft, or third parties.
  • No Improvement Usage: Your data isn’t used to refine OpenAI models or other services.

Security Measures

Your data is protected via Azure OpenAI’s robust framework:

  • Azure Hosting in Your Tenant: The Azure OpenAI service and file uploads are set up in your Azure subscription/tenant, hosted within Microsoft’s secure Azure environment.
  • No OpenAI Interaction: The service operates independently of OpenAI-operated services.
  • Exclusive Fine-tuned Models: If you choose to fine-tune models, they’re reserved exclusively for your use.

For any privacy or security concerns, our support team is ready to assist.

Tracking and Storing Information

In addition to the measures provided by Azure OpenAI, Cetegra AI Chat tracks and stores specific data in our protected database to enhance functionality and provide better insights:

  • Message Tracking: We track the number of messages per bot and the number of tokens (characters) per message. This information helps generate useful insight reports and manage usage limits set by the bot managers.
  • User Access Management: We track which bots different users have access to in order to handle user management efficiently.
  • Chat History: For bots with search history enabled, all messages are stored in our protected database.
  • Uploaded Documents: Documents and data sets you upload are stored in Azure Storage Blobs and indexed. They are only accessible to users with access to the bot associated with the data set. Ensure you control user access and the uploaded documents to prevent unauthorized access.

Key Elements from Azure OpenAI Privacy Policy

Data Processing and Storage

  • Isolation: Your prompts (inputs) and completions (outputs), embeddings, and training data:

    • Are not available to other customers.
    • Are not used to improve OpenAI models.
    • Are not accessible to OpenAI.

  • Data Use: Your data is strictly used to provide the service and to monitor for uses that violate applicable product terms.

Hosting and Interaction

  • Azure Hosting: Microsoft hosts OpenAI models in its Azure environment, and the Azure OpenAI Service operates independently of other OpenAI services like ChatGPT or OpenAI API.

Security Measures

  • Transport Encryption: Uses TLS 1.2 or higher to protect data in transit.
  • Exclusive Access: Compute security measures ensure that your data remains logically isolated with exclusive access to allocated GPU memory.

Data Privacy Guarantees

  • Fine-tuned Models: Any fine-tuned models created using your data are exclusive to your use and can be deleted by you at any time.
  • Augmented Prompts: When using data sources connected to the service, the prompt is augmented to produce grounded generations. Data remains in the designated source location.

Data Storage Features

  • Double Encryption: Data can be encrypted at rest with AES-256 encryption and optionally with a customer-managed key.
  • Deletion: You can delete data stored in the service at any time.

Abuse Monitoring

  • Content Filtering: Real-time evaluation of prompts and completions for harmful content types.
  • Abuse Monitoring Opt-out: Managed customers may apply to opt-out of abuse monitoring. Approved customers can ensure no prompts or completions are stored for abuse monitoring.

Learn More

For detailed information on Microsoft’s data protection policies and compliance offerings, refer to the Data, privacy, and security for Azure OpenAI Service.

By adhering to these data privacy and security measures, Cetegra AI Chat ensures the protection of your data and the privacy of your interactions.

← Groups
Discovery integration →