A decade of open data: Progress, challenges, and how institutions can support

Author: Ed Gerstner

Director of Research Environment Alliances at Springer Nature

For research organisations working to strengthen open science practices, the tenth anniversary of the FAIR principles is a timely milestone to take stock of what has changed and where more efforts are needed. As open data policies evolve, AI adoption accelerates, and expectations for data quality intensify, understanding researcher attitudes has never been more important.

The 2025 State of Open Data webinar brought together leading experts to discuss a decade of findings on researchers’ experience of, and attitudes towards, data sharing. In this blog, I explore key themes and persistent challenges to open data and highlight insights that can help institutions more effectively support meaningful and sustainable data sharing.

Closing the gap between Open Data ambition and everyday practice

Open data is no longer an emerging idea or an aspirational principle. The 2025 State of Open Data report, ‘A decade of progress and challenges,’ analyses ten years of tracking researcher experience and attitudes towards data sharing. It shows that open data has moved from the margins of scholarly practice into the mainstream.

Indeed, awareness is high, policies are widespread, and infrastructure exists. But data sharing is still not routine for many researchers, between mandate fatigue and a lack of recognition.

To examine the tensions between the progress and persistent obstacles, the webinar convened experts contributing to the 2025 report. Their perspectives provide a focused assessment of how far the community has come and what is required next to close the gap between intent and implementation, particularly for organisations aiming to support equitable and sustainable data‑sharing practices.

Webinar panel:

• Mark Hahnel, Vice President Open Research, Digital Science
• Graham Smith, Director Research Data Innovation, Springer Nature
• Brian Nosek, Co-founder and Executive Director, Center for Open Science; Professor of Psychology, University of Virginia (USA)
• Hilary Hanahoe, Secretary General, Research Data Alliance (RDA)
• Joy Owango, Founding Director, Training Centre in Communication (TCC Africa) (Kenya)
• Chair: Ana Van Gulick, Head of Customer Engagement, Digital Science

“Open data has gone from novelty to mainstream… now everyone recognises that this is something that they need to wrestle with.”

- Brian Nosek, Co-founder and Executive Director, Center for Open Science; Professor of Psychology, University of Virginia

From awareness to infrastructure: The decade’s biggest shift

The State of Open Data survey has been running since 2016, capturing researchers’ relationships with data sharing. A decade on, it has impressive scale and global reach. Since its launch, the survey has gathered insights from over 43,000 researchers across 212 countries and territories, making it the longest-running longitudinal study on this topic. It also covers a wide range of institutional contexts, including universities, hospitals, industry, and government.

The webinar asked panellists how researchers’ perspectives on data sharing have evolved over this decade. One of the most significant changes is the steady rise in awareness of the FAIR principles of findability, accessibility, interoperability, and reusability. Mark Hahnel describes this as a genuine success story: “About 80% of researchers globally have heard of FAIR or are familiar with FAIR. ‘Never heard of FAIR’ has gone from around 60% to 20%.” Importantly, this shift cuts across most disciplines, even if adoption still varies.

The deeper shift, as identified by Joy Owango, is the move “from awareness to infrastructure”. Ten years ago, many researchers simply did not know what open data meant or where to share their data. Today, FAIR-aligned repositories, persistent identifiers, and institutional policies are far more common.

The conditions to move beyond counting datasets towards thinking about quality are already here, according to Hahnel. “No one can say, ‘I don’t know where to put my data’ anymore,” he says.

Researchers are often judged by publications and journal prestige, not by the accessibility or reusability of their data sets.”

- Joy Owango, Founding Director, Training Centre in Communication (TCC Africa)

Making data sharing routine: How organisations can support researchers

Awareness and infrastructure have improved so dramatically, but data sharing is still not routine. The problem is not principle, but practice, according to the webinar panel. What opportunities for organisations can we identify in the friction points researchers face that the panel discussed?

1. Helping researchers start earlier can make all the difference: Brian Nosek describes a familiar scenario: Researchers confront data-sharing requirements at the point of publication, long after data collection has concluded. At that point, preparing data for sharing becomes a burden. The well-intentioned policies turn into bureaucratic obstacles rather than good research practice. Organisations can certainly support researchers in prioritising data sharing from the start and guiding them throughout the research process. This support can make sharing an essential part of communicating research rather than an afterthought.
2. Supporting recognition for data sharing in research assessment: Deeper structural issues, particularly around incentives, also play a part. Researchers, Joy Owango notes, are still evaluated on publication and journal prestige, not on the accessibility or reusability of their data. Because of this lack of recognition in funding or promotion, data sharing remains a low priority even where infrastructure exists. Organisations can support the move towards broader recognition of data quality and openness in their own contexts, as well as more broadly in coordinated action and alignment across funders, publishers, and policy frameworks. By providing clear guidance, infrastructure, and training, they can help ensure data contributions are recognised as important research outputs.
3. Capacity differences show where targeted support can have real impac: Hilary Hanahoe highlights the cost of implementing FAIR properly and the disparities this creates: “Some institutions can afford specialised staff to support researchers; others cannot.” Indeed, institutional capacity is uneven, and these gaps are often amplified by regional and regulatory constraints. Where limited capacities are identified, institutions can attempt to address this collaboratively, for instance using joint training programmes or shared guidance as pooled capacity across institutions in a consortium or a region. These adjustments could produce meaningful gains and create more equitable conditions across the research ecosystem. 

AI and open data: Accelerating discovery, amplifying risk

The webinar speakers are in agreement that AI will redefine both the value and the risks associated with open data, and I couldn’t agree more. Adoption of AI tools by researchers is accelerating rapidly, Mark Hahnel notes, and at this pace, AI-assisted research will soon be near-universal. This creates new opportunities to enhance discovery and efficiency, while also highlighting areas where data practices can continue to strengthen.

Because AI systems are based on the input they are given, Joy Owango warns of poor metadata, biases, and mislabelling that can be amplified at scale. Continued investment in robust data stewardship is therefore increasingly valuable, to ensure that data sets are well-documented, clearly structured, and accurately labelled.

AI can also unintentionally narrow the landscape of research by privileging data that is already machine-readable and well documented, according to Graham Smith. That bias risks marginalising disciplines, regions, and forms of knowledge that are harder to standardise. This raises important questions for research organisations about how AI may shape what is visible in global research, and how existing support mechanisms might be adapted over time to avoid further marginalising less structured forms of knowledge.

Looking ahead, Hilary Hanahoe declares that the open data community has a duty, not just an opportunity, to shape the governance of AI use in research. In this important role, the open data community can contribute frameworks that prioritise transparency, inclusivity, and responsible use, and ensure that AI strengthens the research ecosystem.

“Policy on its own isn’t going to do the job. We’re talking about behavioural change, cultural change.”

- Graham Smith, Director Research Data Innovation, Springer Nature

From sharing data to doing it well: How institutions can support better data sharing

The challenge for the next decade is the shift from compliance to quality, from volume to value. And policies and mandates along cannot drive sustainable, meaningful data sharing practices. To support continued and better adoption of data sharing, institutions must increasingly be involved in shaping how research itself is produced, shared, and reused in the age of AI. Not just through their policies, but the actual experience of their researchers and how usable their data is.

As we’ve seen, awareness of data sharing is at an all-time high. This allows institutions to move beyond advocacy to operationalisation. To encourage the necessary cultural change, enabling good practice and making it easier than bad practice is key. Institutions can promote this, where possible, through practical support for data sharing, encouragement of data sharing research practices and discourses, and formal or informal recognition.

In this transitional timepoint, institutional infrastructure, services, and training should aim to support researchers in producing data that is usable and trustworthy. Discipline-specific guidance could round your offerings for holistic data sharing support. This support would ideally be offered early in the research lifecycle as well as throughout, and paired with practical tools such as templates, storage, curation time, and technical support. With this institutional leadership and support, the academic community will continue moving into a future in which data is reusable and reproducible, and the ecosystem is inclusive and trustworthy.

Understanding the state of open data can empower institutions to support data sharing proactively and appropriately. Watch the full webinar for more details and in-depth exchanges.

Related content:

• The 2025 State of Open Data report: Can technology push openness forward?
• Supporting open science: Exploring barriers and solutions to data sharing
• Supporting global open data through what really works
• The many dimensions of rethinking research assessment
• Increasing equity through open access agreements

Don't miss the latest news & blogs, subscribe to The Link Alerts!