Open Science and a coming realignment
Recent developments signal a change in the governance and agenda-setting of Open Science. As the USA withdraws from alliances and positions of leadership, China’s influence continues to emerge.
The USA steps back
UNESCO’s Recommendation on Open Science is reference for those engaged in promoting Open Science and developing policy to support it. Since it was drafted and published in November 2021, it has become a fixed point that can be used to orient discussions about Open Science and the common ground and benefits it can bring.
The United States of America has announced that it will leave UNESCO at the end of 2026. In fact, the USA has signalled its intention to withdraw from a raft of United Nations agencies over the next year. This appears to be part of a reassessment of the country’s global role, and part of a wider pattern of stepping back from alliances and collaborative mechanisms.
This does not equate to a repudiation of the Recommendation of Open Science, but the withdrawal from UNESCO will likely be significant for Open Science. It will impact the USA’s engagement and ability to influence further discussions, and it comes in parallel to an announcement made very late in 2025 by the Chinese Ministry of Science and Technology. Entitled ‘Action Plan for International Cooperation in Open Science’.
The timing highlights a divergence in national policy choices, but also implications for how coordination in Open Science will be approached.
China offers an alternative approach
The announcement is short at under 1,000 words and I would encourage everyone to read it.
The announcement and the objectives set out as a Global South-led1 framework, although China’s position financing and scale means it occupies a central role. The framework suggests how Open Science can be strengthened through international co-operation. It sets out six objectives: governance coordination; data sharing; infrastructure and resources; open journal circulation; public engagement and understanding; and capacity building in developing countries. The Action Plan follows the launch of the Initiative for International Cooperation on Open Science in late 2024 by an alliance of China, Brazil, South Africa and the African Union.
Both the Action Plan and the Initiative prominently reference UNESCO’s Recommendation on Open Science, as well as implementation of the United Nations Sustainable Development Goals. The extent to which they do align with the Recommendation will be interesting, but the signal alongside the US announcement is stark.
These developments should be considered alongside the Open-for-Science 10 (OS10) initiative, a China-led Open Science initiative, launched in 2025, which is coordinated by the Chinese Academy of Sciences (CAS) closely affiliated to UNESCO2.
At the Open Science Fair in September 2025 in Geneva, CAS proposed building a ‘Global Open Science Information Infrastructure Sharing Network Alliance’. This model implies a federated architecture, with interoperability managed across participating infrastructures rather than through a single central platform.
China’s Knowledge & Data Sharing Platform would serve as a core technical anchor within this network; coordination and integration supported by a secretariat hosted by CAS and the China Association for Science and Technology–UN Consultative Committee on Open Science and Global Partnership (CCOS).
The presentation lists, among others Crossref, Open Science Framework and European Open Science Cloud (EOSC) as potential partners. The Knowledge & Data Sharing Platform consists of several Chinese Open Science infrastructures – those named include PubScholar and data centres.
Reconfiguring Open Science
Chinese geopolitical strategy is not easy to read as an outside observer with no special insight into Chinese research policy or industrial strategy; it is easy to misunderstand intentions, and to equally under- and over-estimate the significance of what one is seeing.
However, current developments will affect the centre of gravity of Open Science; as the USA signals its intent to step back, China is offering new approaches to collaboration and shared infrastructure.
The USA’s withdrawal from UNESCO does not mean that it has given up entirely on Open Science. It has been a strong supporter of Open Science and there is a concentration of providers of Open infrastructure, critical datasets and repositories in the USA. While Open Science infrastructure has grown elsewhere, most notably in Europe, this has given the USA the ability to set agendas within Open Science, despite governance being dispersed, informal and international.
It is not clear how much longer this will be the case; the extent of the changes will depend on how USA’s domestic political trends play out, and how the current administration’s formal policy towards Open Science develops. But there will be change, as other countries and agents react to the signal the USA is sending and the uncertainty generated by the changed position.
Discussions about China in research and publishing, at least, those to which I am party, normally centre on its role in driving growth in article volumes and publishing revenues, its emergence as a leader in STEM disciplines, and in relation to diversity and equity in representation in governance and editorial workflows.
The initiatives highlighted here indicate that China is setting out a framework to lead and influence the development of Open Science. This will happen over many years, and will be grounded in the infrastructures that support Open Science. The Action Plan is a call to action; OS10 sets out part of a framework through which this can be implemented.
Changing terms of engagement
The terms of engagement are likely to change for organisations based in the Global North and which provide infrastructure to support Open Science. Debate may shift to whether and how to engage with these Chinese initiatives. What will it mean for an organisation to be a ‘node’ in a global information infrastructure anchored in a Chinese-led programme? Will it remain possible to be a global infrastructure provider, or rather a regional entity plugged into a global infrastructure?
The reference to EOSC’s is particularly interesting, because EOSC itself works through a structure of nodes, although these are intended to be national or thematic units working within its framework. Coordination happens through shared rules, standards and interoperability, with governance set up by the EOSC Association. Will EOSC be able to participate as a ‘node’ itself in this framework? Or will political concerns about EU engagement with China render this impossible?
For policymakers and national governments, including the EU, the question is probably not whether to engage, but rather the extent to which they will. The Action Plan is explicitly flexible and non-binding, allowing partners to judge the most appropriate extent and mechanisms through which they participate. This space for compromise and staged engagement will allow potential partners to manage issues of trust that arise from a China-led scheme of this type.
Crucially, this allows for broad and varied coalition building, and will allow partners to engage while protecting against dependency. This is precisely the type of mechanism that many will be seeking as they feel the impact of the change in approach by the USA. The result will be a net increase in collaboration with China on Open Science and its critical infrastructures, and thus a growth in Chinese influence.
This seems to be a high-stakes opportunity and test for the Principles of Open Scholarly Infrastructure (POSI). POSI 2.0 was released in October 2025 and continues to emphasise governance norms, interoperability, sustainability and the importance of resilience. The initiatives that arise from the Action Plan and OS10 seem well-aligned with POSI. They will be a critical test of both POSI’s relevance, and the extent to which its standards on governance models and interoperability can translate into mechanisms to establish trust and collaboration.
Any Open Science realignment will implications for research integrity and for the sustained growth in publication volumes over time. These issues are not a primary focus of the Action Plan or OS10, but they are nonetheless likely to be shaped by their impact.
As a final thought, it is worth returning to a previous post on the ‘China cycle’ and its relevance for scholarly publishing3. The Action Plan and OS10 are not explicitly about commercial publishing, or journal ownership. They are about a great deal more than publishing: cooperation through integrated platforms, data spaces, applications and standards will shift the anchor of the infrastructure layer underpinning publishing. While this does not amount to the realisation of the China Cycle in scholarly publishing, it is certainly another strong point of reference.
It will be interesting to watch which countries and entities engage with the Action Plan and Information Infrastructure. The pattern and pace of this engagement will be a key indicator of whether signals of intent lead to the realignment of Open Science governance.
Acknowledgement - I’d like to thank Niamh O’Connor, Chief Publishing Officer at PLOS for thought provoking discussions on this topic and signposting to relevant materials.
The term ‘Global South’ is sometimes used synonymously with the term ‘developing countries’, but emerged in part to aid countries in the southern hemisphere work in collaboration. The term itself is used in the announcement.
The specifics of this relationship are not entirely clear - it is listed as a Chinese initiative in some publications or UNESCO-led in others.
In short: China Cycle, as articulated by economist Noah Smith as applied to academic publishing, suggests that China’s research output could see it displace multinationals globally. The first stage of growth with major international partners is clearly underway, but later stages involving global publishing dominance and reshaping market structures are yet to emerge. If Chinese-owned journals, editorial influence, infrastructure, and alternative metrics continue to grow and displace existing models and incumbents, that would represent the realisation of the China Cycle.