MPAS Guidelines for Integrating Generative Artificial Intelligence

The integration of generative artificial intelligence (genAI) functionality is expected to become increasingly prevalent both through enhancements to existing digital tools and through new applications that can enhance both our educational experience and clinical practice. This will likely result in a more hybridized generation of content, blending human creativity with genAI capabilities, thereby reshaping how we approach all aspects of the teaching and learning process as well as the ways in which we practice medicine.

The focus for genAI integration in the MPAS program is to augment, rather than replace aspects of the educational experience we offer to students. Learning is meaning-making, a fundamentally social process with human interaction at the core.

As genAI tools evolve, they hold significant potential to improve learning processes. It is imperative that we remain proactive and agile in leveraging these opportunities. However, this must be balanced with a thorough understanding of the associated challenges to ethical behaviour and be sure to implement appropriate mitigation strategies. Addressing these challenges necessitates the establishment of pillars of trust within our academic and professional communities. These pillars include:

• Generative AI (GenAI): A category of artificial intelligence technology that creates new content (text, images, audio, etc.) in response to prompts. Examples include Microsoft Copilot, Gemini, and ChatGPT.
• Academic Integrity: The commitment to honesty, trust, fairness, respect, and responsibility in all academic pursuits. It involves doing one's own work and properly acknowledging the contributions of others, including genAI tools.
• Transparency: The practice of being open and honest about the use of genAI. In the context of the MPAS program, this means clearly disclosing when and how genAI tools were used to assist the creation of academic work.
• The AI Assessment Scale: A framework used to define five levels of acceptable genAI use, ranging from No Use Permitted to Full Use. These guidelines adopt this scale to provide clear guidance for students and instructors.

USask’s glossary of AI-related terms for robust definitions of these and dozens of other relevant terms.

The following guidelines are meant to contextualize USask’s Artificial Intelligence Principles for the MPAS program.

• Ethical and Effective Use: All use of genAI tools must enhance learning, research, and/or scholarly work while preserving core academic principles. These guidelines are intended to foster excellence, innovation, and inclusivity within the MPAS program.
• Augmentation, Not Replacement: genAI should be a tool to augment, but not replace, the judgment and critical thinking skills of learners, faculty, and staff. As per the College of Physicians and Surgeons of Saskatchewan, students will be held accountable for their clinical decisions, even when genAI is employed.
• Patient-Centered Care: The use of genAI must prioritize patient well-being and safety – including privacy and the right to refuse healthcare informed by genAI. Doing so upholds the essential humanistic aspects of medical practice.
• Foundational Knowledge: In cases where automaticity of core skills or knowledge is essential, genAI will be restricted to ensure foundational knowledge is established in working memory – a critical component of professional competence.
• Transparency and Academic Integrity: All use of genAI must be conducted with integrity and transparency, in alignment with USask’s Academic Integrity policies.

USask’s Guidelines for genAI Use along with the USask genAI libguide provides extensive information to support educators, students, researchers, and administrators in implementing genAI responsibly and ethically. Please consult the guidelines specific to your role(s) in MPAS.

Specific levels of genAI use are differentially appropriate depending upon the learning activities planned for courses. They are based on the principle that some tasks require exclusively human intelligence while others benefit from human-genAI collaboration. The Tiers of Acceptable genAI use outlined below draw heavily from a blog post by Gwenna Moss Centre for Teaching and Learning. In any case, it is recommended that if students are unclear about the tier of genAI use that is acceptable in a particular instance, they should seek clarification from their instructors.

• Level 1:
  • Description: GenAI is strictly prohibited. This level is reserved for assessments that measure core foundational knowledge and skills that must be developed and demonstrated independently.
  • Examples: Certain quizzes, in-class exams, personal reflections, and/or clinical skills assessments where automaticity and recall are the primary learning objectives.
• Level 2: Acknowledged Planning and Scaffolding Use
  • Description: GenAI may be used for pre-task activities such as brainstorming, generating outlines, or performing initial research, but the final product must be the student’s original work.
  • Examples: Using genAI to create a list of differential diagnoses to consider before drafting a patient case study, or to create a structured outline for a paper. 

• Level 3: Acknowledged Editing and Refinement Use
  • Description: GenAI may be used to identify grammatical, spelling, and syntax errors, or to suggest improvements to structure or phrasing. The core content and ideas must still be the student’s own.
  • Examples: Using a genAI tool to proofread a patient chart note (without disclosing patient identity) or a research paper for clarity and grammar.
• Level 4: Acknowledged Collaborative Use
  • Description: GenAI may be used to assist with completing the task, including generating ideas, drafting the paper, and providing feedback. Students must critically evaluate and modify all AI-generated content and transparently disclose the collaboration.
  • Examples: An assignment where students are required to use genAI to draft a patient education handout and then critically edit and improve it based on evidence-based medicine principles.
• Level 5: Full Use
  • Description: GenAI can be used extensively, and in a way that is acknowledged and transparent, to achieve specific learning goals. The focus is on the student’s ability to direct the AI and demonstrate critical thinking about its output. Nevertheless, befitting ongoing professional and legal considerations, we must stress that unacknowledged full use is considered Academic Misconduct – subjecting the individual to USask Academic Misconduct Policies and Procedures.

• Following External Policies: When on clinical learning experiences and/or placements, students must adhere to the specific genAI policies and guidelines of the clinical site or health authority where they are working.
• Privacy and Confidentiality: Students must never input confidential patient information or protected health data into a genAI tool.

Program Manager, Master of Physician Assistant Studies (MPAS) Program
Phone: 306-966-4252