Elsevier Medical Education presents an interview with Dr Ruth Paterson, Associate Professor and Head of Nursing at Edinburgh Napier University, Chair of Innovation in Education Strategic Policy Group at Council of Deans of Health, which discusses responsible adoption of generative AI in nursing education, balancing benefits with ethics, staff support, and equitable access while strengthening critical appraisal skills.
In the report, Principles for the use of Generative AI in healthcare education, for which you wrote the foreword in February 2026, it is noted that 88% of higher education students were already using generative AI in 2025. In what ways do you believe this "bottom-up" reality will shape, or perhaps challenge, the AI adoption strategies within academic institutions?
Bottom-up AI adoption in academic institutions
Well, I think bottom-up approaches are really the cornerstone of quality improvement in both healthcare and healthcare education. As higher education institutions, we’re really student-centered and underpin our practice by a co-design and co-development of programs and our work. One of the things that we had to acknowledge when we started to put together these principles is that a large proportion of students are already using AI, but they may not be using it in a way which informs their practice as well as it possibly could do. But we also know that the research tells us that if AI is used ethically, it has a broad range of benefits. We really need to work with our students to oversee and support them to look at the strengths, the efficiencies and the limitations of AI. Also, it’s an acknowledgement that AI is all around us. Whenever we open up our smartphones, whenever we do internet searches, AI is now integrated into this. Whether we feel that’s the right thing or not, it’s something that is a reality of our life now.
When we started off with the research to come up with these guidelines, we consulted with our Council of Dean members who were cautious about its widespread use. The initial sense and reaction about this two years ago was that Higher Education Institutions (HEIs) should be policing its use and should take a punitive approach in order to preserve the academic integrity of our work through the non-use of AI. I think there’s been a shift over the time and we’ve understood through our institutions that AI literary is really a core attribute for our graduates and we need to look at how we can embed it in a way which is supportive to our students and understands how students, qualified healthcare practitioners, and our service users are using AI.
Responsibly balancing AI’s benefits against its pitfalls
One reporter has reported that banning AI or not using AI is analogous to a pre-electricity-free world. I think that’s possibly a little bit extreme, but it is something that we do need to think about. We need to make it work for us. We need to see the positive benefits while also looking to see what pitfalls there may be.
So really, in terms of my institution, we have underpinned our policy and strategy by research, ensuring any directives are ethical and inclusive in their design. We can’t be over-reliant on AI because this could set challenges around environmental sustainability, student and staff motivation, assessment credibility, and what people’s jobs might look like in the future.
There’s a lot of discussion now about whether or not this means that jobs are going to disappear, and I think in some sectors – including some sectors of healthcare – that might be the case. But I’m a nurse and I think we all appreciate the human side of everything that we do.
Mitigating the risks to student well-being
Finally, we need to mitigate against the challenges widespread AI adoption may present. A recent paper in Frontiers in Psychology highlighted this potential impact on student well-being related to isolation, reduction of face-to-face teaching, and insecurities about what future job prospects are. These might not be new challenges that our students have, but it’s something that highlights that we need to be really careful about how we approach AI in healthcare education and how we can support our students to understand it more.
The report acknowledges that AI tools evolve extremely rapidly, almost on a weekly basis. How should academic staff be supported to ensure they remain up-to-date with these constant technological advances?
I think there are three main areas that we need to think about. One of those is about peer support – and that’s internal and external. I think we need to look to our institutions to support our development in our understanding of AI and thirdly, to also be part of the policy that underpins the use of AI in healthcare education.
Peer support through interprofessional shared learning
So, in terms of peer support, I think it’s an interprofessional approach that’s needed. Shared learning is really fundamental. There are already some good examples of this – we have had a number of excellent conferences run by Public Services Scotland and our AI report that we wrote with the Council of Deans presented a number of case studies to illustrate how we’re using AI in healthcare education.
Institutional investment in AI: financial and time
Thirdly, a lot of institutions have invested in support for AI – for example, Vice Principals for artificial intelligence and AI hubs for Continuing Professional Development (CPD). Also, education of their staff. This is fundamental to staff support and their development. It’s about academics understanding of how AI is actually going to be embedded and how we work with our students to maximize its potential. From a CPD point of view, institutions need to think about protected time for staff to experiment and reflect on the use of AI. They need space to test the tools in teaching context, not only hear about them in theory.
Communities for AI support and practice
In healthcare context, that’s also to test and think about how we can really develop the understanding of these tools. As for these communities of practice – early adopters will always be there. People that are real champions of AI. Then there’ll be those that are more hesitant and there’ll be everyone else in between. With those who are more hesitant, we need a forum for listening about concerns, finding out about best practice, and understanding the transparency, all of which were key principles in our report.
"We need to treat AI as a support tool that shapes or supports professional judgment. And we need to teach our students, like we do with everything else in our healthcare practice, that it’s fundamental to verify that any AI-generated information is legitimate and will be for the patient’s benefit."
Dr Ruth Paterson, Associate Professor and Head of Nursing, Edinburgh Napier University | Elsevier Faculty Hub, June 2026
You mentioned that AI adoption must balance innovation with ethical considerations. In your view, what is currently the greatest ethical challenge facing healthcare educators regarding generative AI?
I think it goes back to the fundamental ethical principles that we all know in healthcare, which is about respect for choice and weighing up benefit, harm, and justice. If I may, I’d say that I think there are two main challenges here. One is about preventing harm, making sure that we do what we can to mitigate risk and harm to our public that we’re treating. Also, justice and making sure that our student population has equitable access to the tools that they can use to develop their understanding.
Challenge 1: preventing harm to patients
So, starting with the prevention of harm. I think back to when I first started using AI – it’s very engaging, it’s very polite, and very credible when it gives you answers, which is very reassuring. There has been lots of work about the AI platforms and the chatbots that are even replacing human relationships. I think we have to be mindful that it’s quite biased in its imagery and indiscriminate in its curation of information. Therefore, AI is a starting point rather than the font of all knowledge. Human factors absolutely need to be central to the development of AI in healthcare education. We need to treat AI as a support tool that shapes or supports professional judgment. And we need to teach our students, like we do with everything else in our healthcare practice, that it’s fundamental to verify that any AI-generated information is legitimate and will be for the patient’s benefit. So, to reduce that harm, I think our students and our staff should be trained to check information curated from AI, identify hallucinations, recognize bias, and make judgements about when AI is inappropriate for a particular task.
Challenge 2: the gap in equitable access
Then, I think a second challenge is about equity. If some students have better access or more AI fluency than others, those students could potentially be more advantaged, and it can widen the gaps in achievement and confidence in healthcare delivery. Whilst it can bridge the gap for people whose first language might not be English or who may have additional support needs, we must also be mindful that it could widen the gap as well.
Regarding equitable access, how can we prevent AI from widening the educational gap for students with fewer resources?
Institutional responsibility for equal access to AI tools
I think we learned a lot from the pandemic and students with challenges regarding access to the internet, particularly in the remote and rural areas of the UK. We need to be mindful as institutions and as a profession to give everybody access to the same tools. Those tools are ones that we can train our staff and our students to use. We can support our students and our staff as much as possible to verify the outputs that they get from these AI platforms and to evaluate. Evaluation of AI is fundamental to promoting student success, without being detrimental to the population that we’re being trained to serve. We need to think about student progression and AI literacy. Certainly in my institution, there’s buy-in about a particular AI tool so that there is an opportunity for everybody to access a uniform supportive AI tool that our students can use.
We’ve got to think about how students access the web. Some will still be using data to access the internet resources that we provide. So, it’s about offering low bandwidth and mobile friendly options where possible. Students may not have a PC or laptop. They’ll use their mobile devices to access information. So, it’s really, really important that whatever we’re offering is universally accessible.
Also, providing support for our students, which we do anyway with any program that we’re running. Drop-ins and peer mentoring, giving students the opportunity and the platform to ask anything about AI and really supporting them in their learning about AI. And then, as I said, monitoring.
Universal designs for learning with AI
These are all the principles of universal design for learning, which is about making sure that whatever we design fits the widest population that we possibly can. We’re a widening access university here at Edinburgh Napier and what we’ve got to make sure is that we’re giving everybody the opportunity to achieve their potential, whatever that looks like, and AI is one way to support that.
"I think there’s real strength in working with AI to improve some of the inefficiencies in health and education systems – and in considering how we can use AI to do some of the tasks that take away from the face-to-face care and support that is so important to not only nursing, but allied health and medicine"
Dr Ruth Paterson, Associate Professor and Head of Nursing, Edinburgh Napier University | Elsevier Faculty Hub, June 2026
As a leader in nursing education, what message would you give to professors and librarians who feel fear or distrust toward using AI in their professional development?
Why skepticism about AI is healthy
I think skepticism is good and it’s okay to feel that way, and it is really important for us all to hear those concerns. We have to understand and respect the professionals who are hesitant in its use. Everybody likes a shiny new thing and there’ll be those who are early adopters of it and there’ll also be people that are very, very hesitant, and there are lots of reasons why people are hesitant in its use. For example, if we look at the sustainability argument against AI, there is a very strong argument about its use and why we shouldn’t be using it. But there’s also an efficiency around its delivery and its support that we also need to think about. As health professionals, we’re there to protect the public in terms of privacy, data sharing, confidentiality, and we need to understand what AI can do to enable that, but also the associated risks. I think we need to take a research approach to this. As academics, we’re all trained in the research process and making sure that we’ve got rigor, transparency, and applicability when using AI is important. So, the first thing is that we need to think about exploring, testing, and building the evidence for its reliable capability.
How AI can support health service users
We need to start experimenting on what the AI platforms are going to give our service users in relation to clinical conditions or medicines information and how verifiable and transparent that information is. I think there’s a real opportunity for AI to summarize information and make it more understandable for service users with health literacy challenges. I’m talking about our learning disability population and our cognitive impairment populations who may be supported in prompts or in understanding how they use their medicines or how they approach their health challenge.
Using AI to cut through research overload
We know that there’s a vast array of health research information that professionals have to sift through to make a judgment on a particular intervention. This can be overwhelming, particularly for our students. Having AI to support some of this work may be of benefit. But we need to encourage people to explore its capacity to do this at their own pace, and this is very different. This is taking a person-centered and a student-centered approach to realize that not everybody’s going to develop their knowledge at the same pace and there’s no uniform way to do this. As academics and healthcare professionals, we have a real responsibility to develop rigor in appraisal of AI and rigor in dissemination of its worth. We need to be part of the AI development as healthcare professionals. That was really central in the report that we wrote. I think there’s real strength in working with AI to improve some of the inefficiencies in health and education systems – and in considering how we can use AI to do some of the tasks that take away from the face-to-face care and support that is so important to not only nursing, but allied health and medicine. This needs to be taken forward and explored using the research skills that we’ve all developed and that we teach our students.
Bringing research rigor to testing AI
We need to test hypotheses in relation to AI. We need to use reproducible methodologies to investigate AI interventions and use precision to inform the direction of future practice in this area. I think if we do that collectively as a collaborative course across the professions with our students, our healthcare users, and with ourselves, this will be a real winner for us in the use of AI.
In your opinion, what is the unique human skill that a future healthcare professional must strengthen most in the face of advancing AI?
I think, as humans, we have a unique ability to take a critical eye – seeing different points of view and trying to eliminate bias. And really, rather than believing everything that we see in AI or that is in AI, we need to appraise it. I think we need to develop structured models to appraise the information provided by AI that protects the public from bias, inappropriate information disclosure, and unreliable information. That is absolutely vital. We then need to think about how we can integrate it into our systems. What’s the practical application of AI into the problem that we’re seeking to address? Whether that’s to develop more robust and up-to-date teaching plans, summarizing the large amounts of clinical information that might be presented to us, or verification of the information the patients have curated – that’s very, very important. Then, we need to evaluate. We need to continue to appraise its use.
So, I would say that the human skills are about asking good questions around AI and how it can support our work, appraising the answers and the outputs that it gives us, and thinking about whether or not that information is reliable enough to integrate into our healthcare systems. And then, do continuous evaluation to ensure that we’ve got a really integrated, important system that sits alongside what we do, which is essential to health professionals for that caring, treatment, and human aspect of healthcare delivery.
Leave a Reply