Appendix A: Framework for AI Adaptation at TMU
Appendix A
Generative Artificial Intelligence Task Force Report
Appendix A: Framework for AI adaptation at TMU
| Strategic Focus | Key Components | Goals & Actions | Outcomes | AI Training | Infrastructure & Tools |
| Adaptation and Innovation in Learning and Teaching | Grounded in existing principles: literacy, ethics, transparency, learner-centred innovation, rigour, equity, accessibility, and discipline connection | Support instructors, students, and departments to make thoughtful, pedagogical AI decisions | A robust repository of effective pedagogical practices and interventions hosted by CELT | Multi-stage support for diverse readiness levels Faculty: Support for assessment redesign, course design, data privacy, and modeling appropriate AI use via CELT, Libraries, and the Chang School Students: Developing ethical decision-making, use, understanding and evaluation (e.g., TMU Libraries' AI Fluency Badges) | Institutionally supported AI tools (Gemini, NotebookLM) Centralized pedagogical repository hosted by the Centre for Excellence in Learning and Teaching (CELT) to scale up and share emerging strategies across the institution Expanded facilities access for invigilated in-person assessments |
| Non-reactive, intentional, and discipline-specific guidance | Offer Designated Teaching Fellowships & Students as Partners in GenAI grants | Active local experimentation grounded in shared institutional values | |||
| Academic freedom and inclusive representation of diverse teaching views | Integrate AI competency directly into curriculum and program design (via PPR, department consultations) | Development of program-level learning outcomes inclusive of AI competencies and skills | |||
| Assessment reform & secure evaluations | Establish program-level "assurance spines" (oral defences, in-class labs, supervised exams) and authentic, creative assessment redesign | Shift from reactive "AI-proofing" to long-term authentic learning & verified student competencies | |||
| Catalyst for Scholarly Research and Creative Activity | Open approach balanced with security, privacy, and ethics | Invest in and support the next generation of AI research talent, innovations, and intellectual property | Breakthrough research and advanced knowledge created in a public-trust-aligned environment | Continuous skill development and talent-cultivation pipelines across the research ecosystem Training in reflexivity, recognizing algorithmic limitations, and implementing secure research data management practices | Shared, multi-model, and scaled platforms across the Ontario university sector High-performance secure compute infrastructure Data sovereignty tools to safeguard IP and sensitive research inputs |
| Academic freedom matched with reflexivity to prevent misuse and protect sensitive data | Leverage AI tools to address complex research questions and solve intractable societal problems | Improved administrative efficiencies, allowing researchers and staff to redirect time/resources to higher-value activities (student support, strategic planning, and partnership development) | |||
| Recognition and minimization of AI tool limitations | Expand access to AI models and tools to support SRC | Broadened research opportunities and discovery potential | |||
| Sector-wide collaboration | Drive the provincial AI innovation ecosystem and participate in shared AI infrastructure initiatives | Leverage shared resources from expanded AI capacity and data sovereignty | |||
| Safe and Secure Operational Efficiencies | Risk management balanced with administrative efficiencies | Create expert-guided, safe experimentation spaces Permit low-risk use cases | Improved operational efficiency and adoption of AI tools to support improved service levels and response times | Expert-guided workshops on prompt engineering and output verification Broad staff training on identifying AI-related risks and championing responsible, safe use Peer-supported prompt libraries and collaborative knowledge-sharing through a university-wide community of practice | Centrally supported tools (e.g., Gemini, NotebookLM) Centralized Use-case Registry and Tool Repository Data governance and classification guidelines AI Incident Response Guide |
| Tiered approval model categorizing tasks by risk level | Enforce formal management review, explicit authorization, and human-in-the-loop oversight for high-risk use cases | Ensure safety and compliance with policies, laws, and regulations | |||
| Collaborative community of practice | Regular opportunities for staff to share and learn about operational uses of AI for efficiency | Build capacity and support AI competencies | |||
| Core Safety Components: Data governance, centralized registry, clear approval chains, vendor assessments, and incident response | Establish staff guidelines for AI use | Clarity of permissible use to enable innovations and efficiencies |