Generative AI for administrative tasks at UBC: What staff need to know
Generative AI (GenAI) is becoming a key part of how we work—across teaching, research, and administration.
This page serves as an entry point to navigate GenAI for administrative tasks at the Faculty of Medicine. It incorporates resources from Digital Solutions, UBC IT, and UBC PrISM (Privacy & Information Security Management).
General Guidelines
Use Cases
GenAI can be used at UBC for administrative tasks such as:
- Drafting emails and communications
- Summarizing meeting notes or reports
- Brainstorming ideas and solutions
Prompt Engineering
To get better results, use structured prompts like the RTF framework:
- Role: Define who the AI is (e.g., “You are a communications assistant…”)
- Task: Specify what you want it to do (e.g., “Draft a follow-up email…”)
- Format: Indicate the desired output (e.g., “In bullet points” or “As a formal memo”)
More resources about prompt engineering on the CTLT site: https://ai.ctlt.ubc.ca/prompt-engineering
Privacy & Security
- Follow UBC’s Principles for the use of GenAI tools to mitigate risks like bias, plagiarism, and data misuse.
- Conduct Privacy Impact Assessments (PIAs) when handling sensitive data or deploying tools broadly.
- Refer to the practical list of Do’s and Don’ts in the next section
Accuracy & Oversight
- Always review and verify AI-generated content before sharing or publishing.
- Treat GenAI as a support tool, not a decision-maker.
Transparency
- Be clear when content is AI-assisted, especially in formal communications.
- Maintain accountability for final outputs.
Practical Privacy & Security
Do’s:
- Use approved AI tools that meet UBC’s privacy and security standards.
- Manage data carefully—only share what’s necessary.
- Limit what you input into AI tools.
- Review AI-generated outputs for accuracy and alignment with UBC policies.
- Engage with privacy and security teams early when planning to use AI.
- Share responsibly—privacy and security is a team effort.
Don’ts:
- Do not enter personal or confidential information into unapproved AI tools.
- Do not reuse your CWL password in any AI service.
- Do not accept vendor terms without reviewing privacy and security clauses.
- Do not assume that anything you input into an AI tool is private.
GenAI Tools at UBC

M365 Copilot Chat is UBC’s GenAI Tool of Choice
Use M365 Copilot Chat (via m365.cloud.microsoft) with your UBC email address.
Practical Use-Cases at UBC:
- Simplified Communication: Quickly draft clear, professional emails or summarize key points from lengthy email threads.
- Meeting Efficiency: Generate succinct meeting summaries, action items, and follow-up communications directly within your workflow.
- Project Support: Receive instant organizational support for managing project tasks, timelines, and collaborative documents.
- Content Creation: Rapidly outline or draft reports, presentations, and instructional materials, significantly reducing preparation time.
Getting Started with Copilot Chat:
Open Copilot Chat and begin leveraging AI-driven insights and assistance immediately.
Important Security and Privacy Guidance:
- Always authenticate using your firstname.lastname@ubc.ca account, this ensures your enterprise data is protected.
- Copilot Chat is approved for handling only low to medium-risk data per UBC’s Information Security Standard (ISS U1)
More info and updates: https://it.ubc.ca/news/june-06-2025/m365ubc-now-copilot-chat
Other GenAI Tools for UBC Work

List of AI tools that have undergone UBC privacy and information security review, with associated risk profiles and use conditions: https://privacymatters.ubc.ca/sites/default/files/2025-10/PIA_Reviewed_AI_2025-09.pdf
DeepSeek is the only restricted tool at UBC. More info here.
Training & Support
UBC IT has general training sessions available for M365, including Copilot: https://it.ubc.ca/microsoft-365/general-training-microsoft-365
UBC is also developing role-specific AI training for staff, faculty, and students. More to come.
Reach out to Digital Solutions for further questions, including support with AI adoption or concerns outside of Administration: dcio.medicine@ubc.ca
FAQ
What can I use for meeting summarization? Can I leave personal or sensitive data?
You can currently use Copilot Chat to summarize meeting transcripts. Meeting transcripts can be obtained from Zoom or MS Teams if the meeting was recorded.
The current version of Microsoft Teams (including Copilot chat) available to all UBC users is approved for Medium Risk data or lower. This means it’s acceptable to use Copilot chat with employee names and to generate transcripts that include staff-related information. However, if your content includes learner information, it must be de-identified before using Copilot chat, due to current licensing and data protection standards.
UBC is actively reviewing and testing a more advanced version of Copilot. If and when this version becomes available—pending licensing agreements—it will support the use of High Risk data, allowing for broader administrative applications.
How do I know what data is safe to use?
UBC Electronic Information is classified as follows:
- Low risk includes names and work contact info of UBC staff & faculty, and anything publicly available, etc.
- Medium risk includes proprietary information, confidential financial info and records, etc.
- High risk includes personal information, including student names, student & employee IDs, face ID photos, grades, home addresses, etc.
- Very high risk includes SIN, official gov ID, bank account info, PHI, DoB, etc.
More info: UBC data classification (ISS U1): https://cio.ubc.ca/information-security-standards/U1
What other AI tools are safe to use for UBC work?
All solutions that have been reviewed by the UBC Privacy & Security office are available here: https://privacymatters.ubc.ca/sites/default/files/2025-10/PIA_Reviewed_AI_2025-09.pdf
DeepSeek is the only restricted tool at UBC. More info here.
My unit has its own guidelines – what should I follow?
Some units, programs and departments may classify certain type of data (e.g. schedules) as higher risk than UBC’s general standards. This is especially true in the Faculty of Medicine, where health-related data is common. In such cases, follow the specific guidelines set by your unit, even if they are more restrictive than UBC-wide policies.
Going Further
Resources for Faculty: https://facdev.med.ubc.ca/artificial-intelligence/
Framework for GenAI in Education: UBC Faculty of Medicine’s principles and recommendations for integrating GenAI into educational practices: https://facdev.med.ubc.ca/resource/ubc-faculty-of-medicine-framework-for-generative-artificial-intelligence-ai-in-education/
AI & Health Network: A collaborative hub advancing AI research and innovation in health at UBC: https://aihealth.med.ubc.ca/
Artificial Intelligence and Technology-Enhanced Care Collaboration Centre (AiTECCC): https://aiteccc.ubc.ca/
A FOM AI Strategy is in development. More information will be provided.
UBC GenAI Training Hub for Staff: Centralized training resources to support staff in using GenAI tools responsibly: https://genai.ubc.ca/training-hub-for-staff/
UBC IT AI Services: Overview of AI-related services and support offered by UBC Information Technology: https://it.ubc.ca/services
UBC VPRI on AI: UBC’s research strategy and initiatives related to artificial intelligence: https://research.ubc.ca/ai
UBC GenAI Guidance Website: Guidance on tool categories, approval pathways, and responsible use: https://genai.ubc.ca/guidance/
Guidelines for AI tools at UBC: https://privacymatters.ubc.ca/sites/default/files/2025-10/PIA_Reviewed_AI_2025-09.pdf
CTLT resources on AI for Teaching and Learning: https://ai.ctlt.ubc.ca
Privacy Matters @ UBC Website: UBC’s central hub for privacy guidance, including PIA guidelines, FAQs, and reports: https://privacymatters.ubc.ca
PIA Guidelines for AI Solutions: Specific guidance on assessing AI-related tools: https://privacymatters.ubc.ca/sites/default/files/2024-09/PIA_Guidelines_GenAI.pdf
UBC Self-Service Portal for PIA Inquiries: Submit and track Privacy Impact Assessment inquiries and requests: https://ubc.service-now.com/selfservice
UBC Information Security Standards (ISS): Defines risk classification for data and systems, access controls, and lifecycle requirements. https://cio.ubc.ca/information-security-standards/U1
UBC Policy SC14: Administrative heads’ responsibilities for information security and acceptable use: https://universitycounsel.ubc.ca/policies/information-systems-policy/
UBC Signing Resolution #26 (Clickthrough Agreements): Defines who may accept online vendor terms on behalf of UBC: https://universitycounsel-2015.sites.olt.ubc.ca/files/2022/07/Resolution-26-Clickthrough-Agreements.pdf
BC Freedom of Information and Protection of Privacy Act (FIPPA): Provincial legislation governing collection, use, and disclosure of personal information: https://www.bclaws.gov.bc.ca/civix/document/id/complete/statreg/96165_00
AI Acknowledgement: This content was produced with limited assistance from Microsoft Copilot.