As part of our Annual Forum this year, we’re taking the opportunity to hear directly from the organisations supporting the event and learn more about their perspectives on Open Science and open research. Through this interview series, we’ll explore what openness means to different stakeholders, the opportunities and challenges they see across the Arab region, how they are supporting researchers and authors, and what they hope to see develop across the research ecosystem.
We spoke with Mark Hahnel, VP of Open Research at Digital Science and Founder of Figshare, about the role of open tools and infrastructure in supporting researchers, the challenges of sustaining open science, and the impact of AI on research dissemination.
Emily Choynowski: Tell me a bit about your organisation, Digital Science.
Mark Hahnel: Digital Science started in 2010 as a spin out from Macmillan Publishers. The aim was to build research tools to make research more efficient and revolutionise the way science is done. As an organisation we believe in the power of research to shape the future, solve global challenges, and drive innovation that benefits everyone. So Digital Science started out with a few investments in startups from members of the research community, one of which was Figshare – which I started out of my PhD at Imperial College London to help with data publishing. We now have anywhere between 12 and 14 products that have been born of real-world challenges and serve all aspects of the research community from the researchers themselves, all the way up to universities, publishers, industry, governments, policy makers, and funders.
Emily: Why does your organisation support open science?
Mark: As VP of Open Research, I see our commitment to openness as part of the organisation’s founding principles. Our vision is that “research is the single most powerful transformational force for the long-term improvement of society”. Digital Science came about as the web began transforming scholarly communication, at a time when open access was gaining momentum and governments and funders were recognising the benefits of making research openly available. Projects like the Human Genome Project demonstrated the scientific and economic value of sharing research, alongside its broader contribution to society and the Sustainable Development Goals (SDGs). Open research also matters in practical terms, because people should not lack access to important research simply because they are not affiliated with a university, especially when that information could affect their lives.
Our leadership has always believed that research should be as open as possible, so others can build on what has come before. But making research reusable requires tools, curation, and services that support researchers’ workflows. These can provide a sustainable business model, while improving equity and inclusion. Since 2010 Digital Science has grown driven by innovation to support every corner of the research ecosystem.
At Figshare, one of our founding principles was that research impact should not depend on where work is published. Research should be “as open as possible, as closed as necessary.” We also encourage the use of Creative Commons licences such as CC BY and CC0. Researchers need to be able to protect their intellectual property where appropriate, but unfortunately many have not received enough guidance on licensing, which can lead them to choose unnecessarily restrictive terms.
Emily: Beyond making research available, how does Digital Science facilitate the growth of open science and the development and implementation of open science policies and practices across different research communities?
Mark: We try to give back to the research community. Our products serve researchers, institutions, publishers, and funders, and we also want data to move freely between systems, avoiding unnecessary lock-in. For example, our solutions address different parts of the research workflow. Figshare provides data and full institutional repositories; Symplectic Elements is a research information management system that allows institutions to capture research outputs automatically and deposit them into their institutional repository through bi-directional integration, helping meet open access policies without adding to researchers’ workloads; Dimensions is a link research database which brings together information on publications, patents, clinical trials, policy documents, datasets, pre-prints, awarded grants and more, to see where open research is happening and which collaborations and funding shape it; and Altmetric tracks online engagement and reveals where research is receiving attention, from social media to policy papers, the press and patents. When you want information to flow freely between all your systems, you are able to see where the gaps are. And so, at Digital Science we identified a need for persistent identifiers for research institutions. We developed GRID and made its database openly available on Figshare, without commercial restrictions. GRID later contributed to the development of ROR, which has become a widely used standard for identifying research organisations.
Also, where we can support the wider ecosystem by sharing data or software, we aim to do so. Overleaf, for example, makes substantial parts of its collaborative writing codebase openly available and has grown into a widely used product with around 30 million users. This shows that open-source development can coexist with a sustainable business model.
Figshare has also released open-source tools developed in collaboration with other repositories to improve interoperability, including through shared metadata standards. Funders need to be able to search across repository infrastructure, so open standards and non-proprietary approaches benefit the wider ecosystem as well as the organisations using these tools.
Emily: What do you see as the biggest barriers to making open science more sustainable, particularly given current pressures on higher education and research funding?
Mark: The world has changed significantly in the past three years, and it’s becoming a little harder for organisations to maintain the same levels of commitment amid changing pressures. I’m optimistic about AI and its potential to accelerate progress on major challenges, including SDGs. But it will also disrupt businesses and change how we think about academia. Subject experts will be increasingly important in evaluating AI-generated work and ensuring that these systems are used responsibly.
The current research system illustrates the principle that incentives shape outcomes. A culture of “publish or perish” and an emphasis on impact factors encourage researchers to publish more. Open access has also created different financial incentives, and some publishers have developed successful models at very different price points. These organisations can publish high-quality research, but the wider question is whether the system is producing too much research, and whether it can assess that output effectively. Annual publication volumes have increased substantially, from around two million in the year 2000, to around eight million last year, and AI will continue to accelerate that trend.
The challenge is whether AI can help curate, filter, and verify new content quickly enough to keep pace with attempts to game the system. We’re already seeing publishers and conferences ask authors to demonstrate that they understand the work they submit. At Figshare, which accepts preprints, we have had to invest more time and resources in reviewing submissions and filtering out pseudoscience. Once content is indexed by search engines and AI systems, unverified claims can be presented as facts, even when they have not undergone peer review.
These pressures are affecting open infrastructure. Platforms face large volumes of unsuitable submissions and automated bots downloading data at scale. Some services have had to change how they operate or change for certain levels of access. This raises difficult questions about what openness means in practice and how we ensure that researchers, including those in the Global South, can access the tools and resources they need.
Emily: What might help address these challenges, particularly for researchers and institutions in the Global South that have limited resources and may struggle to protect their platforms from AI bots crawling and overwhelming their systems?
Mark: I think we need to develop a more nuanced understanding of what openness means. When the FAIR principles were introduced, I was primarily concerned with making research findable, accessible, interoperable, and reusable for people. We now need to distinguish between openness for people and openness for machines. These uses do not necessarily have the same needs or impact. We should consider how to label and distinguish human-generated, AI-assisted, AI-generated, and human-curated content, and decide what expectations and safeguards should apply to each.
There is also a significant cost to servers from large-scale automated downloading by AI systems. That activity should not be treated in the same way as a researcher in the Global South accessing material to conduct research and build on existing knowledge. We need approaches that preserve meaningful access for people while addressing the pressures created by automated systems.
Emily: How do you see the role of publishers, funders, librarians, technology solution providers, and other stakeholders evolving in the coming years as open science continues to develop?
Mark: I think we all need to focus more on how research is made available and useful to different audiences. APIs have already enabled people to build new services and applications on top of research in ways that were previously difficult to imagine. But now, I think that every organisation will need to consider how to give researchers a good experience when publishing their work, ensure they receive appropriate credit, and make research easy for people to discover and read.
We all also need to consider machine users and how research can be made accessible to them responsibly. AI could help optimise research tools and reduce repetitive administrative work. Researchers spend a great deal of time on grant applications and reformatting manuscripts for different journals, for example. Better tools could remove some of this routine work and give researchers more time to focus on producing high-quality research.
Emily: Is there anything else you would like to share with our audience about Figshare or Digital Science more broadly?
Mark: I think we were fortunate to enter the research technology space when we did; we came from the community and built many of our tools with open principles in mind because we care about the people, the research itself, and what universities and other users need. Figshare.com as a free platform for individual researchers to share their own research outputs according to the FAIR principles remains available. We provide the storage and mint the DOIs to continue encouraging open sharing. At the same time, sustainability requires a financial model and that’s why Digital Science is a commercial business. Keeping everything free and open is not always straightforward – services require ongoing maintenance, hosting, information security, support services and investment in development. Working on a sustainable model helps ensure that tools remain reliable and useful over time, which provides credibility and assurance to our users. We’ve spent 15 years building the systems, processes, and certifications needed to work with organisations from universities to government agencies. That experience helps us continue to develop new tools while applying principles of open research and open science. As VP of Open Research, an important part of my role is to identify ways to make research more open and encourage those we work with to adopt more open practices. I think that’s an important commitment for a commercial organisation.
