Note: If you haven't already, please also see DoIT’s AI Meeting Assistant Guidance website, created in collaboration with the Office of Compliance and Office of Legal Affairs, to review university key points, unit-specific direction, and personal considerations with this technology. 

1. If you meet virtually with students for class or other purposes, be aware that AI meeting assistant tools are becoming more common and can be enabled by meeting hosts (instructors) and participants (students). To protect data privacy and security, the only AI meeting assistants that should be used at UW–Madison are those within Zoom and Webex services provided by UW-Madison. Consider whether or not it is helpful to use an AI meeting assistant in various teaching settings. For example, you might decide it’s helpful for a study session to capture notes that can be shared with the class but not for meetings that might discourage open conversation. For your class sessions, consider the risks and benefits of having AI generate a summary or transcript and the need to review it for accuracy and maintain it as a public record.

2. Be aware that lecture materials and recordings are protected intellectual property at UW–Madison. Students must have your permission to record lectures (including with an AI meeting assistant) unless they are considered by the university to be a qualified student with a disability who has an approved accommodation that includes recording. If a student is recording a class without your permission and without an accommodation, you may tell them to stop and discuss with you later any needs they may have.

3. Be transparent with students about whether and how an AI meeting assistant tool will be used during class sessions or other meetings. Explain the “why” behind your decision.

4. When you are the meeting host, you control whether an AI meeting assistant can be used by you or by a student. When scheduling meetings, use security settings in Zoom or Webex to allow or restrict AI meeting assistant use. Make sure you are using the most current version of your meeting software.

5. If an AI summary or transcript is shared, assume it is now a public record. As the meeting host, you are responsible for ensuring that it is reviewed for accuracy and corrected before it is shared and stored according to the appropriate records retention schedule.

6. AI tools such as transcription are great universal design strategies. They may make meetings more accessible to both disabled and non-disabled individuals, but they do not necessarily meet all accommodation needs, particularly when professional services are needed. It is important to be clear with students about using any of these AI tools before the class or meeting; refer them to the accommodation process if they need different assistance to access the course or meeting due to a disability. Consult with the McBurney Disability Resource Center if needed.