Human–artificial intelligence (H-AI) collaboration: An emerging role in healthcare
Authors: Dr Frank Kiwanuka1,2, Dr Joanne Smith-Young2,3, Kip Bonnell2.
1. University of Eastern Finland
2. Memorial University/University of Toronto
3. JBI Memorial University Nursing for Evidence-Based Healthcare
In the past three years, since the launch of ChatGPT, there has been substantial recognition that artificial intelligence (AI) systems offer new access to knowledge for patients, families, and healthcare professionals. By July 2025, there were approximately 700 million weekly active users of ChatGPT alone. This tells us why AI matters now. People are already using it to ask questions, seek support, and make sense of health information. The question is how to ensure AI remains human-centred, evidence-based, and responsive to the needs of patients and families.
Keeping Humans in the Loop
AI is rapidly advancing in healthcare, including nursing science. Keeping humans ‘in the loop’ ensures it enhances – rather than replaces – clinical judgement while respecting patients’ diverse beliefs, values, and preferences.
Human-in-the-loop (HITL) is a framework in which AI serves as a supervised collaborator, helping to prevent automated bias from replacing clinical expertise. As a result, final diagnostic and treatment decisions remain firmly in the hands of healthcare professionals.
Given that nurses make up the largest segment of the global healthcare workforce and spend the most time with patients and their families, their role in the design and deployment of AI systems in healthcare is essential and cannot be overlooked.
In a recent letter in Ovid, a leading medical research platform, blog co-author Kiwanuka, and Shafik, propose that AI tools alone cannot guarantee better outcomes. In line with HITL principles, patients and healthcare professionals are central, offering the opportunity to integrate AI tools with societal values and clinical care. These stakeholders are uniquely positioned to translate AI-generated outputs into meaningful actions that align with patients’ needs, preferences, and values.
As health professionals face increasing use of AI tools, their application brings both opportunities and responsibilities for safe and equitable practice. In that context, Kiwanuka and Shafik advocate for a holistic, patient-centred approach that bridges the gap between abstract AI technologies and compassionate, actionable care – ensuring that AI remains human-centred.
Negotiating, Adapting, and Integrating AI
In a recent study on stroke care, Kiwanuka et al. (2025) found that although AI-driven interventions improved diagnostics, treatment, and rehabilitation, their effectiveness and acceptance depended strongly on patient and family perspectives, trust, and communication.
One finding stood out to Kiwanuka – while he initially expected strong evidence of AI’s potential to improve risk detection and enable earlier stroke intervention, he instead observed how often the patient and family voice was absent from the development and implementation of AI tools across the stroke care trajectory.
This realisation reinforced an important lesson – AI tools can enhance patient- and family-centred care when they are developed and used in partnership with patients, families, and healthcare professionals. Importantly, when those stakeholders co-design these tools and are engaged in feedback loops, their experiences and insights can support meaningful participation in care decisions.
At Memorial University’s Faculty of Nursing, Derrick Walsh is an emergency nurse practitioner (NP) and educator with 23 years of frontline clinical experience and five years teaching in NP education. His work focuses on innovating in teaching, strengthening clinical decision‑making, and integrating AI to enhance learning, support growing enrolment, and prepare students for evolving practice.
He leads hands-on workshops in advanced health assessment, casting, and suturing, and regularly presents nationally on advanced practice topics. At the 2026 Atlantic Region – Canadian Association of Schools of Nursing (ARCASN) conference, he presented Teaching Smarter, Not Longer, demonstrating how AI can support teaching efficiency while maintaining educational quality.
A central theme of Walsh’s presentation is improving productivity without compromising clinical or educational standards. AI tools – such as automated summaries, content generation, and task identification enable educators to ‘teach smarter, not longer’, freeing time for higher-value teaching activities.
Crucially, this approach reinforces an HITL model: AI supports decision-making, but educators and clinicians remain responsible for interpretation, context, and application.
Walsh also highlights a ‘Case Study to Podcast’ approach. In this workflow, clinical cases are converted into structured scripts and then into short podcast-style learning segments. Using simple URL-based links embedded within presentation slides, learners can move seamlessly from case content to audio outputs.
This approach reflects the growing importance of flexible learning formats. Walsh notes that podcast-style content can be created to emphasise different competencies and support students in patient care. For example, a podcast can focus on developing a differential diagnosis, planning non-pharmacological management, or creating follow-up plans. This format allows learners and healthcare professionals to engage with the material during commutes, at the gym, between shifts, or during short periods of downtime, making learning more accessible without sacrificing depth.
Together, these strategies demonstrate how AI can enhance – not replace – teaching and clinical reasoning, while supporting learner-centred education.
These insights also have broader implications. They can help inform the design of AI-enabled tools – such as predictive models or decision-support systems – that are grounded in patient experience and real-world practice.
Nursing Leadership
In their scoping review, Kiwanuka et al. (2025) highlight the growing importance of nurse leaders in guiding AI adoption for workforce planning, staffing optimisation, workload management, and resource allocation. Achieving this requires leadership that combines technological literacy with evidence-based healthcare (EBHC) management. The authors emphasise that nurse leaders should critically evaluate AI tools and assess their impact on care quality and workforce well-being, ensuring decisions are grounded in evidence rather than enthusiasm for new technologies.
One of the most promising applications of AI is in supporting patients and families throughout complex care journeys. This is especially important during critical illness, where uncertainty, emotional distress, and long-term health impacts affect both patients and their loved ones. The human–AI (H-AI) interaction perspective is an emerging paradigm in this space that emphasises the interplay between humans and AI-driven decision-making systems.
Responding to the Needs of a Population Facing Unique Challenges
Our JBI-affiliated group, JBI Memorial University Nursing for Evidence-Based Healthcare, highlights the growing role of nursing scholars in translating AI-generated evidence into education and clinical practice. We aim to lead in AI-supported workforce upskilling and to advocate for patient- and family-centred AI across complex care trajectories.
In Newfoundland and Labrador – Canada’s easternmost province, with a population of just over half a million – our health system faces unique challenges related to distance, access, and equity. These realities remind us that sustainable AI solutions must be grounded in the voices, experiences, and partnerships of the communities they aim to serve. Patients, families, and communities in rural and remote regions bring valuable lived experiences that can help define the problems AI should address and shape the design of AI-enabled EBHC systems. Their voices are essential to ensuring that AI innovations are culturally appropriate and practical in real-world settings.
Reflecting EBHC Pillar 1 – positioning communities, patients, and the public as partners –researchers at Memorial University are helping shape AI approaches locally and globally. For instance, they are using AI to improve prosthetics, detect disease through movement analysis, and optimise healthcare systems, such as emergency room wait times.
Memorial University is also recognised as a leader in Atlantic Canada in the broader field of computer science, which includes software development and engineering, AI and machine learning, and cybersecurity and networks, according to the World University Rankings by Subject 2026. In response to growing demand for AI expertise and the anticipated rapid expansion of this field in the coming decades, Memorial University now offers a Master of Artificial Intelligence program.
Our Future Focus
As we celebrate World EBHC Day 2026, we view AI not as a replacement for healthcare professionals but as a tool to strengthen evidence-based decision-making – one that should be held to the same rigorous standards of evidence that guide clinical care and health policy.
The future of EBHC lies in integrating technological innovation with compassion, empathy, and meaningful partnerships with patients and families. The real challenge is ensuring this technology is balanced with strong clinical evidence, ethical considerations, and a commitment to human-centred care.
Lessons Learned
Based on our learnings, patients and communities must help shape AI across the full care trajectory – from conception and design to testing, implementation, and continuous monitoring. Their voices are essential to ensure AI tools are accessible, reliable, ethical, and grounded in real-world care.
AI-enabled EBHC must involve communities, patients, and the public not simply as end users, but as partners. Their lived experience helps define the problems AI is needed to solve – ensuring innovation addresses real needs rather than theoretical ones.
Ultimately, sustainable AI-driven health systems depend on combining technological innovation with community knowledge and oversight. By embedding patient and public voices throughout, we can ensure AI improves care in ways that reflect the values, priorities, and lived realities of the communities it serves.
References
IBM Technology. (2026, March 17). What is human in the loop with AI? How HITL shapes AI systems [video]. YouTube. https://www.youtube.com/watch?v=9iS-YYLIXiw
Kiwanuka, F., & Shafik, W. (2025). Nursing as a lingua franca for artificial intelligence in patients’ care trajectories: Where are we headed? Nursing Research, 74(3), 170. https://doi.org/10.1097/NNR.0000000000000799
Kiwanuka, F., Shafik, W., & Babirye, M. E. (2025). State of the art of artificial intelligence in patient and family-centered care during critical illnesses: The stroke care trajectory as an exemplar. International Journal of Caring Sciences, 18(3), 1684. https://www.internationaljournalofcaringsciences.org/docs/47.kiwanuka.pdf
Kiwanuka, F., Stevanin, S., Ahtisham, Y., Owusu, B., Nurmeksela, A., & Kvist, T. (2025). Nurse leadership and artificial intelligence integration in nursing workforce management: A scoping review. Journal of Advanced Nursing, 82(6), 5675–5686. https://doi.org/10.1111/jan.70296
Memorial University (2024, December 19). Memorial University ranked No. 1 nationally for using AI to enhance and accelerate research. https://www.mun.ca/marcomm/news-articles/memorial-university-ranked-no-1-nationally-for-using-ai-to-enhance-and-accelerate-research.php
Memorial University (2026). Master of Artificial Intelligence. https://www.mun.ca/become/graduate/programs-and-courses/artificial-intelligence/
Sigalos, M. (2025, August 4). OpenAI’s ChatGPT to hit 700 million weekly users, up 4x from last year. CNBC. https://www.cnbc.com/2025/08/04/openai-chatgpt-700-million-users.html
Times Higher Education (2026). Computer science world university rankings 2026. https://www.timeshighereducation.com/world-university-rankings/2026/subject-ranking/computer-science
Walsh, D. (2026, June 18–19). Teaching smarter, not longer: An integrated Al framework for NP education [presentation]. Atlantic Region – Canadian Association of Schools of Nursing (ARCASN) Conference 2026, St. John's, NL, Canada. https://acrobat.adobe.com/id/urn:aaid:sc:US:8929b16d-3150-4ba4-911c-610dabc773c3
To link to this article - DOI: https://doi.org/10.70253/DHBD7913
Disclaimer
The views expressed in this World EBHC Day Blog, as well as any errors or omissions, are the sole responsibility of the author and do not represent the views of the World EBHC Day Steering Committee, Official Partners or Sponsors; nor does it imply endorsement by the aforementioned parties.
