In the rapidly evolving landscape of higher education, artificial intelligence (AI) tools like ChatGPT are transforming how students approach learning, particularly in demanding fields such as health professions. A groundbreaking qualitative study from the University of Ottawa has shed light on this shift, documenting widespread adoption among health sciences students and underscoring the pressing need for formal training to harness AI responsibly.
Health professions education encompasses training for careers in medicine, nursing, pharmacy, physiotherapy, and related disciplines, where precision, ethical decision-making, and continuous learning are paramount. As generative AI—large language models capable of producing human-like text—gains traction, educators and institutions face the challenge of integrating these tools without compromising academic integrity or clinical competency.
Key Insights from the University of Ottawa Research
Published in JMIR Medical Education on April 2, 2026, the study titled "Artificial Intelligence in Health Professions Education: Qualitative Study of Student Experiences" involved 51 students from 10 health professions programs at the University of Ottawa. Participants, predominantly women (80%) aged 20-29 (76%), represented fields like health sciences (53%), medicine (16%), and psychology (16%).
Through semistructured interviews and an online survey conducted in 2024, researchers identified ChatGPT as the dominant tool, used by 96% of respondents. Other mentions included Microsoft Copilot, Perplexity AI, and Anthropic Claude, but none rivaled OpenAI's flagship model in popularity. Students described AI as a complementary aid rather than a replacement for traditional study methods.
Daily Applications: From Summarizing to Skill-Building
Students leveraged ChatGPT for diverse tasks, enhancing knowledge acquisition and skill development. Common uses included:
- Summarizing research papers and clinical guidelines (55% of participants).
- Generating sample exam questions and rubrics for self-assessment (45%).
- Refining writing and data analysis (43%).
- Clarifying complex concepts and brainstorming problem-solving approaches (31%).
"I have used ChatGPT to help summarize papers and generate great sample questions and to read my work and compare it to the rubric," shared one participant. Another noted, "ChatGPT is good at helping me refine my data analysis. It helps me distinguish between certain nuances." These anecdotes highlight AI's role in streamlining repetitive tasks, freeing time for deeper critical thinking.
Nearly 59% deemed AI essential for modern learning, citing efficiency gains amid heavy workloads. Curiosity (sparked in late 2022) and peer recommendations drove initial adoption, with trial-and-error as the primary learning method.
Critical Evaluation: The Double-Edged Sword of AI Outputs
While enthusiastic, students emphasized vigilance. Over 57% routinely cross-referenced AI-generated content with peer-reviewed sources or personal knowledge to mitigate inaccuracies and biases. "I compare it to peer-reviewed articles to see if the findings match up," one student explained.
Essential skills identified included technical proficiency (43%), critical thinking (37%), and digital literacy (33%). Barriers like unreliable outputs and over-reliance risks were acknowledged, reinforcing AI's status as a supportive tool, not an authoritative one.
Student-Led Recommendations for AI Training
The study revealed a consensus on the need for structured guidance. Preferred strategies encompassed:
- Hands-on demonstrations and workshops (51%).
- Peer-to-peer learning sessions (37%).
- Accessible online resources and tutorials (31%).
- Integration into group projects for practical exposure (25%).
"Concrete exercises would be more useful. For example, using it as part of a specific group project," suggested a respondent. This call aligns with experiential learning theories, advocating for curricula that build adaptability and ethical AI use.
Canadian Medical Students Echo Similar Trends
Beyond Ottawa, a multicenter survey across six Ontario medical schools (167 respondents) found 79% of students using generative AI, with 53% doing so weekly. ChatGPT led, applied for reviewing content, guideline summaries, and differential diagnoses. Notably, 93% expressed willingness to train for clinical integration, though 92% recognized inaccuracy risks and 79% biases. Some 76% supported formal curriculum inclusion.Read the full multicenter survey.
Evolving Policies at Canadian Universities
Institutions are responding swiftly. The University of Ottawa's Faculty of Medicine policy governs AI/large language model (LLM) use in research and writing, promoting transparency and prohibiting unauthorized substitutions for original work. Broader university resources include ChatGPT FAQs and tailored educational versions.University of Ottawa AI Policy.
Other Canadian universities, from Queen's to Toronto, balance bans on misconduct with task forces rethinking integrity. This patchwork signals a sector-wide pivot toward guided integration.
Emerging AI Training Initiatives in Canada
Forward-thinking programs are addressing the training gap:
- McMaster's "AI in Health Professions Education" course explores practical applications.
- Michener Institute's AI in Health Care Certificate focuses on data literacy and ethics.
- Unity Health Toronto's Health AI Academy boosts clinician AI literacy nationwide.
- Waterloo's WatSPEED AI program targets public health professionals.
These initiatives emphasize governance, ethics, and hands-on skills, mirroring student calls from the Ottawa study.
Challenges: Ethics, Equity, and Limitations
Despite promise, hurdles persist. The Ottawa study's single-site focus and self-reported data limit generalizability; larger, multi-institutional research is needed. Ethical concerns—plagiarism risks, equity in access, and over-dependence—demand robust safeguards. In clinical contexts, AI hallucinations could mislead future practitioners without training.
Stakeholders, including faculty and administrators, must foster multi-perspective dialogues to ensure equitable benefits.
Photo by The 77 Human Needs System on Unsplash
Future Outlook: AI as a Pillar of Health Education
Looking ahead, AI could revolutionize simulations, personalized tutoring, and lifelong learning in Canadian health professions programs. By 2030, expect standardized curricula embedding AI literacy, potentially improving outcomes like diagnostic accuracy and research efficiency.
For universities, proactive investment in training positions them as leaders. Students, equipped with critical skills, will enter a healthcare workforce where AI augments human expertise, enhancing patient care.
This Ottawa study serves as a clarion call: embrace AI thoughtfully to empower the next generation of health professionals.