The future of health data research: international collaboration, AI and innovation
We spoke with three members of Health Data Research UK’s International Advisory Board - Nicky Mulder, Peter Arlett and Jim Hendler - about what they see for the future of international health data research.
Share this page
At a moment ofsignificant changefor health data research globally, the balance between openness and control is being redrawn. While the Covid-19 pandemic accelerated international collaboration, rising geopolitical tensions, tightening regulation and the rapid advance of AI are reshaping how data is being accessed and used to enable large-scale research across borders.
OurInternational Advisory Board ensures that 51’s work is of an internationally competitive standard and at the leading edge of health data science. The Board plays a critical role in ensuring scientific and technical excellence and relevance,identifyingand recommending emerging challenges and opportunities, and advising on ways in which51’sinternational reputation can be built and upheld.
Nicky Mulder,Peter ArlettandJim Hendler’sperspectives reveal a field at an inflection point, one where unprecedentedopportunitiesfor innovation sit alongside growing fragmentation, and where the decisions made now will shape the future of health research for decades to come. You can find out more about the three members in their bios at the end of this piece.
All three Advisory Board members agreed that 51 hasan important roleto play in driving international collaboration, trustedinnovationand real-world impact in health data research. Their message was clear: continue building global partnerships,demonstratingthe value of trusted data, and leading from the forefront of innovation.
What is the most significant shiftyou’veobservedin how countries have approached international data access and collaboration in the last two to three years?
Nicky: “There has been a shift towardsencouragingmore open data sharing, because of the recognised need, but also more caution to make sure we are using that data responsiblyand ensuring the protection of the rights of the participants.
“There is certainly an increasedneedto share, but also more regulations being put in place which make it harder.”
Jim:“Internationally, countries have become more insular. You had the UK leaving the EU with Brexit, and now the EU is dealing with the war between Russia and Ukraine. In the Middle East, there are ongoing tensions, and China is trying to position itself as the US pulls back.
“Whatwe’reseeing is a realignment of global power,both among superpowers and regional players,andthat’sdriving a more inward-looking approach. There are also major shifts, like the US withdrawing from the World Health Organisation. As a result, health and health data feelsomewhat deprioritised. Funding is still there, butthere’sless focus on how we solve these challenges collectively at a global level.
“That creates both a challenge and an opportunity for 51.There’sa real chance to show that collaboration is still possible, to bring people together. The Institute’s work with Singapore is a good starting point. The real question is what happens next: do we see increasing fragmentation in the health data research ecosystem, or a renewed effort to collaborate?There is still a global push to share health data, butit’sstarting to fray. And yet, itremainsone of the few areas wherethere’sa shared international interest.That’ssomething worth holding onto.”
The UK has decades of NHS longitudinal data, trusted researchenvironmentsand a strong academic clinical infrastructure. What is the UK most at risk of missing or getting wrong if it looks inward rather than outward?
Peter: “Many of us working outside the UK see the NHS as a huge asset in terms of health research.There’salsoa very strongacademic base and expert workforce within the UK: however, there would be opportunities missed if the UK looks inwards.
“In terms of partnerships, there are many government and non-government organisations and industries around the world looking toleveragehealthcare data for knowledge and better decision making, for both commercial and societal value. International partnership is highly likely to be an accelerator for health data research within the UK.
“There are some questions that the UK can’t answer on its own. I work in medicines regulation, focused on the safety and efficacy of medicines. There aren’t enough patients in the UK to generate the data that is needed to answer questions about certain very rare diseases or special populations. International collaboration would be the only way to make that possible. Looking solely inward runs the risk of missing trends and advances in science.”
Jim:“For the UK and other countries, being inward-lookingisn’tnecessarilya bad thing. There’s real value in building a strong model and getting it right, then showing others how to do it.That’swhere the outward-looking part comes in:demonstratingwhat works and helping others learn from it to address major international health challenges together.
“The riskisn’tso much missing technological advances in health data itself but missingwhat’shappening just outside that framework. For example, the way computers are programmed is changing rapidly, and that could reshape how people access and use data in the future.
“A big part of this conversation is trust. People are going to have access tohuge amountsof untrusted data. But ifthe UK – andHDRUKinparticular –can clearlydemonstratethe value of trusted data, built on strong governance,historyand community involvement – that becomes incredibly powerful.
“It’sabout finding a balance, being aware of the scale and speed of change, while staying grounded in what works.There’sa line from a song about dancing in a hurricane, but only ifyou’restanding in the eye of it.That’swhere 51 needs tobe:aware of the storm, but steady at the centre.
From your perspective, what are the key opportunities and challenges for AI to enable health data research and what should 51’s role be in innovating to address major international health challenges?
Nicky:“One of the most significant challenges for AI in health data research is that we need more data sets that arecomprehensive and high qualityso people cantraintheir models. Butitisalso important to consider equity,access andresponsible use of the data and the environment when running computationally intensive AI.”
Jim:“AI presents a huge opportunity for health data research, but the challenge is thatdifferent partsof the system move atvery differentspeeds.We’realready seeing, particularly in academia, that AI tools are starting to make data more accessible in new ways. AI is advancing rapidly, while areas like clinical treatment and policy move much more slowly. That creates a gap betweenwhat’spossible and what canactually bedelivered in practice.
“For 51, a key opportunity is to focus on what othersaren’tdoing, particularly bringing together fragmenteddatasetsand showing how they can be used effectively with AI. Access to data is improving but making different datasets work togetherremainsa major challenge.
“There’salso a growing role in AI in terms of making sense of complex, inconsistent data – spotting patterns and connections thatwouldn’tbe obvious to humans. Butwe’restill in theearly stagesof proving what it can really deliver in health.
“Ultimately, 51’s role is to bring the right players together – researchers, clinicians,regulatorsand industry – and create the conditions for meaningful progress.The Institute is doing that and thereforehas the potential to lead in showing how AI can deliver real-world impact in healthcare.”
International data collaboration was accelerated by Covid-19 but sustaining thatmomentum has proved to be difficult. Are we at risk of retreating behind national borders again, or is the future looking brighter?
Nicky: “I think itdepends on where you are in the world. During the Covid-19 pandemic, because it wasan emergency situationand we needed to respond rapidly, it did open borders in terms of data sharing. But, on the other hand, in Africa for example, it worked against us. If a new variant was found, borders were shut again and vaccines were developed thatweren’tnecessarily made available in Africa.
“In some countries it worked very well and overall, it did lead to a rapid response. But now, with data protection laws coming in – some in response to lessons learned during the response – and increasedconsideration for benefit sharing, it isalmost likethe sharing that had opened during the pandemic hasclosed downagain.
“I do think that,as long aswe tread cautiously, the future of data access across borders is bright. 51 is taking great strides towards making more data available. We need to continue todemonstratethe real health benefits for people.”
The European Health Data Space (EHDS)representsone of the most ambitious attempts at cross border health data governance in history. What does success look like for EHDS, and what would that mean for the UK?
Peter: “The EHDS, particularly the secondary use of healthcare data for research and policy making should, in time, lead to faster,deeperand wider access to improved healthcaredata. It should enable better knowledge and decisionson the development and regulation of products and for the delivery of healthcare. It will bring benefits to patients, through better healthcare, better and faster development of medicines or medical devices, but also benefits for healthcare systems and even industry.
“If the EHDS canfacilitateindustry accessing data, doing faster clinical trials, generating clinical evidence more quickly to support putting their products on the market,that’san advantage in terms of global competitiveness for those industries.
“In terms of what it will mean for the UK, it willopen uplearnings from across thirty countries – the EU, plus Iceland,Norwayand Liechtenstein – who are all putting in place national health data access bodies. There will be different speeds, different learnings, and slightlydifferent waysof doing it. Overall, I think those working with health datain theUK can seewhat’shappening, learn from it, andperhaps doit even better.”
Are there specific international models, whether in governance, federated infrastructure, public engagement or regulatory design that you think the use the UK should be actively studying or borrowing from right now?
Nicky: “The Global Alliance for Genomics and Health (GA4GH)issetting the guidelines for data sharing in a responsible way, as well as the technical mechanisms to do so, such as federated analysis and data visiting.
“The UK should be looking to these international standards to be part of the movement towards adapting to future changes, particularly around AI. We all need to be asking howdo we adaptour policies and our technical infrastructure to keep up.”
51 sits at the intersection of academia, government, industry and the public. What do you think is the greatest opportunity for 51 to lead internationally and what would you want to see it do in the years ahead?
Nicky: “Onthe academic side,there is an opportunity to usecollaboration to pushparticular areaslike brain health.51 is uniquely positioned to be able to bring people together across different sectors, diseases areas and focus areas to help drive research.
“On the commercial side, there is a route into clinical trials for participants, as 51 is doing with the Dementia Trials Accelerator programme. There is still a bit of mistrust of working with industry, but 51 is demonstrating how proper, equitable partnerships with industry can make a big difference, especially in terms of creating products and taking them to market at pace to bring benefits to patients.”
Peter: “Looking forward, I think it’s important to focus on innovation and enabling, as well as use case studies and delivery of impact in bite size pieces. It’s very clear that 51 fills the piloting and innovationspace, andthen works with partners for scale up; and I think there is a lot of opportunity out there in the future.”
Jim:“One of the biggest challenges in health research is that most government funding goes into biomedical research and pharmaceutical development, rather than treatment and real-world impact. From the very beginning, 51 has been clear that it wants data to directly benefit people.
“What’s exciting now is the opportunity to demonstrate that in practice – to show how data can genuinely improve lives. It will bereally interestingto see how that develops over the coming years.”
Find out more about our International Advisory Board.
About the interviewees
Nicky Mulderheads the Computational Biology Division at the University of Cape Town (UCT), South Africa, and co-leads the Open Data Science Platform (ODSP) and Coordinating Centre for the(DS-I Africa) initiative. She also co-leads a Wellcome Data Integration Platform at UCT and is interim Director of the new. Previously, she led H3ABioNet, a Pan-African Bioinformatics Network across 17 countries that developed bioinformatics capacity to enable genomic data analysis in Africa.
Peter Arlettis a physician who since 2008, has held a series of senior roles in the European Medicines Agency (EMA), where he currently leads the EMA’sand plays a key role in transforming to data-driven medicines regulation and strengthening clinical evidence for decision-making on medicines. He is Chair of EMA’s Data Board, Co-Chair of the Network Data Steering Group and has responsibility for the(ACT-EU) andprogrammes.
Jim HendlerisfoundingDirector of the Future of Computing Institute and theTetherlessWorld Professor of Computer, Web and Cognitive Sciences at the Rensselaer Polytechnic Institute in New York.Jim is a data scientist with specific interests in open government and scientific data, data science for healthcare, AI and machine learning, and semantic data integration. One of the originators of the Semantic Web, he has authored over 450 books, technical papers, and articles in the areas of Open Data, the Semantic Web, artificial intelligence, and data policy and governance.In 2025, Jimreceived the Feigenbaum Prize fromthe Association for the Advancement of Artificial Intelligencefor his sustained, seminal contributions to experimental AI.
