Canvas M-Connector for Faculty and Students

Bring Canvas course information into U-M GPT

The Canvas M-Connector helps U-M GPT find and use authorized Canvas course information.  Faculty and instructors can use U-M GPT to review course content and activity, grade work, message students, and update course materials, with approval required before any changes are made.  Students can ask it to list assignments due across courses, identify missing or late work, check grades and submissions, summarize discussions, and explain topics using supported course materials.

The M-Connector experience is designed to support teaching and learning while preserving faculty control. Access to full course materials is managed independently for each Canvas course.

Learn about all M-Connectors

Choose your path:

For faculty and instructors

You decide how the Canvas M-Connector fits your course

The Canvas M-Connector is enabled at the course level. Faculty and instructors control whether the Canvas M-Connector is available only to instructors or to both instructors and students.

When course-material access is enabled, students can ask U-M GPT questions about supported content from that course. When it is not enabled, full course materials remain unavailable through the connector.

What faculty can do

Depending on your Canvas role and permissions, the connector can help you:

  • Review course content and structure: courses, syllabi, pages, modules and module items, files, assignments, announcements, discussions and replies, and rubrics.
  • Find or summarize teaching materials: search page/file names, inspect course structure and content, read supported course files, and summarize course organization.
  • Review student/instructor workflows: upcoming assignments, the user’s submissions and grades, assignment submissions and analytics, student engagement analytics, and extensive peer-review status/completion/quality reporting.
  • Create course materials: pages, modules and module items, assignments, announcements, discussion topics, and rubrics.
  • Update content/settings: supported settings and content for pages, modules, module items, assignments, discussions, and announcements.
  • Canvas messaging: draft content in chat and send supported Canvas Inbox messages, including bulk and peer-review reminders. Multi-recipient sends require a preview/confirmation workflow.
  • Prepare a course: inspect and reorganize existing materials, upload files, copy course content through a migration workflow, and create or update supported content.

U-M GPT works with your existing Canvas permissions; connecting it does not grant additional access. Before U-M GPT changes Canvas content, sends a message, submits work, grades, or takes another supported action, it asks you to review and approve the action. U-M GPT cannot complete any action without your confirmation.

How faculty get started

  1. Click M-Connectors in the right-side navigation. 
  2. Click Enable. 
  3. Click to choose if the Canvas connector should be available only to you or to everyone with access to your course.

Refer to the Turn On the Canvas M-Connector in Canvas knowledge article for detailed instructions. 

Prompts faculty can try

  • “Using Canvas, give me an overview of the structure and published content in my [course name] course.”
  • “List the upcoming assignments in [course name] and flag scheduling conflicts or unusually busy weeks.”
  • “Review the organization of my [course name] modules and suggest improvements. Do not change anything.”
  • “Draft an announcement reminding students about this week’s deadlines. Show me the draft before posting it.”
  • “Scan the supported course content in [course name] for basic accessibility issues and give me a summary. Do not make changes.”
  • “Help me organize materials for a new course using Canvas content I can already access.”

Recommendations for faculty

  • Start read-only. Ask U-M GPT to find, summarize, compare, or draft before asking it to make changes.
  • Tell students what is enabled. Explain whether the course has the Canvas M-Connector enabled, what information students can use through U-M GPT, and how this aligns with your course AI policy.
  • Review generated work. Verify summaries, dates, grades, messages, and proposed changes against Canvas before relying on them.
  • Keep a human in the loop. Use U-M GPT to prepare consequential work, then review it carefully before approving any action.
  • Choose the right experience. If you need a curated student-facing tutor aligned with defined learning outcomes and selected course materials, consider Canvas Maizey instead of, or alongside, the general-purpose Canvas M-Connector. Maizey Agentic Chat app includes the ability to add M-Connectors to your Maizey project. 

For students

What the Canvas M-Connector means for you

The Canvas M-Connector helps you use U-M GPT with Canvas information you already have permission to see. It can help bring together information that might otherwise require checking several Canvas pages or courses.

Your instructor decides, course by course, whether U-M GPT can use the full supported course materials. If an instructor enables that access, you can ask questions about supported course content. If the instructor does not enable it, those full course materials remain unavailable through the connector.

What students may be able to use in an M-Connected course

When an instructor makes supported course materials available, students may also be able to:

  • Assignments and deadlines: See upcoming work across courses and identify what is due next.
  • Grades and submissions: Check grades, feedback, and whether an assignment was submitted, is missing, or is late.
  • Study planning: Turn Canvas to-do items and deadlines into a study plan without changing anything in Canvas.
  • Assignment guidance: Summarize instructions and identify questions for clarification.
  • Discussions and messages: Review course conversations and catch up on important communications.
  • Course materials: Find supported content, summarize readings, or explain a topic using accessible course sources.

What students can access when full course materials are not enabled

Connect the Canvas connector in U-M GPT  to organize your work even when faculty haven’t enabled the Canvas connector. Once active, ask U-M GPT: “Which of my Canvas courses can you access, and what are my upcoming assignments.” Due dates will be included with a list of your assignments. 

How to tell which courses are M-Connected

Log into U‑M GPT and ask, “Which of my Canvas courses can you access, and what can you tell me about my [course name]  assignments?” If full course content is unavailable, the Canvas connector has not been enabled for that course.

How students get started

Connect the Canvas M-Connector, enable the connector for your conversation, then ask U-M GPT about your courses, upcoming assignments, to-do items, or submission status. Or, start by testing out the prompts below. 

Prompts students can try

  • “Using Canvas, list my assignments due in the next two weeks, grouped by course.”
  • “Show my Canvas to-do items and help me make a study plan. Do not change anything.”
  • “Which of my Canvas courses can you access, and what information is available from each?”
  • “What is my submission status for [assignment] in [course]?”
  • “Summarize the instructions for [assignment] and identify anything I should clarify with my instructor.”
  • “Using the materials available in [course], explain [topic] and cite the course sources you used.”

Canvas M-Connector or Canvas Maizey?

Both connect AI experiences with Canvas, but they support different goals.

Choose the Canvas M-Connector when you want

  • A general-purpose assistant in U-M GPT that works across Canvas and other connected services.
  • A tool that assists with finding, summarizing, organizing, drafting, or taking supported actions based on existing permissions.
  • An experience initiated by the individual faculty member or student through conversation with U-M GPT.
  • Flexibility that combines Canvas context with services such as Google Workspace, Slack, Zoom, or Dropbox.

Choose Canvas Maizey when you want

  • A course-specific AI tutor that students access inside the Canvas course.
  • Greater faculty control over the selected data sources, system prompt, behavior, and intended learning experience.
  • An experience explicitly designed around course learning outcomes or a defined instructional purpose.
  • A shared, curated AI experience for students rather than each student’s general U-M GPT conversation.

Academic Canvas Maizey projects do not require a Shortcode. Students open the Maizey tutor inside the course and complete a short authorization the first time they use it.

Learn about the U-M Canvas Maizey Integration

Get started with the Maizey Canvas Connector

Privacy, permissions, and responsible use

  • The Canvas M-Connector works within your existing Canvas permissions. It does not provide access to courses or content you could not otherwise access.
  • Course-material access is managed course by course.
  • U-M GPT asks you to review and approve supported actions before they are completed.
  • AI-generated answers can be incomplete or incorrect. Verify important information in Canvas and do not replace human judgment with AI.