Skip to main content

AI & Connectors

What Are Connectors?

Connectors are tools that let GenAI platforms (like ChatGPT and Gemini) securely link to your external services, such as email, cloud storage, or calendars. They boost productivity by letting AI work with your existing apps and data.

However, they also come with privacy considerations you should understand before turning them on.

Tip
Please review this page before enabling ChatGPT or Gemini connectors. You will find instructions for enabling connectors below in the "Instructions for CES Users" section.

Connector Features

Chat Search
Quick, one-off lookups across an app. Ideal for everyday tasks like "Find last week's roadxmp in Box" or "Look up the latest version of the Team OKRs in OneDrive." Results appear inline with links.
Deep Research
Broad, multi-query analysis across numerous sources (internal + web), returning cited answers. Examples include incident retros, competitive analysis, code reviews, etc.
Synced*
Proactively indexes selected files so ChatGPT can answer questions without re-querying the source speeding responses and improving quality.

*Currently, Google Drive only

Examples of Popular Connectors

  • Google Drive Connector: Lets the AI tool access and retrieve files or folders from your Google Drive
  • Outlook Connector: Connects the AI tool to your Outlook email and calendar
  • Microsoft Teams Connector: Allows the AI tool to interact with Teams messages, channels, and files
  • GitHub Connector: Enables the AI tool to search, read, and summarize code repositories
  • Other Supported Connectors: Additional integrations are available, such as linking with cloud storage or productivity apps

Approved Connectors

The following connectors have been approved and enabled:
    ChatGPT
    Box, GitHub, Google Calendar, Google Drive, Microsoft OneDrive, Outlook Calendar, Outlook Email, SharePoint, and Microsoft Teams.
      Gemini
      Google Drive, Google Docs, Google Keep, and Google Tasks.

      Why Privacy Matters with Connectors

      When you connect external services, you’re granting AI access to potentially sensitive information. Even if the AI doesn’t store the data, it can process, summarize, or act on it. That means:

      • You should know exactly what data is being accessed
      • You should have a clear purpose for connecting it
      • You should review and adhere to the CES Privacy Principles while using AI

      A summary table of these principles and examples related to connectors is below. You can also learn more about these privacy principles in the links below.

      Principle Example
      Purpose Limitation Before enabling or leveraging a connector, determine the specific purpose for accessing data.
      Data Minimization Don't pull data into an AI tool via a connector that is not tied to a specific purpose and that you do not have an appropriate and legal justification to process.
      Lawfulness Don't violate any privacy and policy agreements. Refer to BYU's Data Use, Privacy, and Security Policy and the CES Information Classification document for more information.
      Transparency Inform users when you use connectors that may access their data.
      Protection Use only services that have been approved. Consult the current approved list.
      Duration Use connectors with retention settings and delete or de-identify data once the task is complete.

      Guidelines for Enabling Connectors

      When you connect external services, you’re granting AI access to potentially sensitive information. Even if the AI doesn’t store the data, it can process, summarize, or act on it. That means:

      • You should avoid enabling connectors with unapproved tools/platforms
      • You should know exactly what data is being accessed
      • You should have a clear purpose for connecting it
      • You should ensure the connector and service provider meet security standards

      Learn More About Privacy Principles and the Responsible Use of AI

      We use AI guided by Integrity, Data Protection, Transparency, and Accountability.
      Learn how to apply the six privacy principles while using AI.
      AI-driven data analytics can unlock insights that drive innovation and improve outcomes. However, this data still needs to be protected