EU turns research with China into a security question, study finds

Between 2021 and 2025, the European Union has quietly rewritten the rules of scientific cooperation with China, reframing knowledge flows that were once presumed open as strategically sensitive resources requiring screening, eligibility tests and, in some cases, outright exclusion. That is the central finding of a qualitative policy analysis published as a research article in Asian Review of Political Economy, in which Yang Li and Ting Long trace how the EU’s “de-risking” agenda has been translated into concrete instruments of research governance.

The study draws on securitisation theory and examines publicly available EU policy documents, including the Horizon Europe regulation and its participation rules under Articles 22(5) and 22(6), the European Economic Security Strategy and the 2024 Council Recommendation on enhancing research security. These are compared with Chinese Ministry of Foreign Affairs readouts and state-media commentaries. In total, the authors analysed 23 public documents: eleven EU regulations, policy documents and official web items, four EU-oriented think tank or specialist media pieces, four Chinese official statements and four Chinese official media commentaries, using an “event window plus keyword search” strategy across 2021 to 2025.

The conceptual core of the paper is “knowledge securitisation”—the process by which research collaboration, data sharing, researcher mobility and academic exchange are recast as objects with security properties, to be monitored, filtered and where necessary restricted. This goes beyond familiar policy labels such as technology de-risking or protection of critical technologies, which target specific fields like semiconductors or quantum technologies. Knowledge securitisation instead describes a deeper shift in governance: knowledge production itself becomes an object of security policy, moving research from “unconditional openness” toward “conditional openness”.

The empirical story begins with Horizon Europe, the EU’s flagship research-funding framework for 2021 to 2027, where most of the rule changes affecting cooperation with China are concentrated. Under its predecessor Horizon 2020, Chinese institutions could participate widely, typically as self-funded partners, and co-funded flagship projects in areas such as climate change, agriculture and food were supported by matching funds from China’s Ministry of Science and Technology. Horizon Europe broadly continued this architecture, with China remaining an important though not legally “associated” partner, most visibly in two co-funded flagships on climate change and biodiversity and on food, agriculture and biotech, in which Chinese participants receive no direct EU funds.

The tightening began with legal groundwork. Articles 22(5) and 22(6) of the Horizon Europe regulation empower the EU, where necessary, to exclude entities from non-associated third countries from innovation actions to safeguard the Union’s strategic assets, interests, autonomy and security. In 2022, the EU and China launched negotiations on a new joint roadmap for science, technology and innovation, but by late 2023 the Commission acknowledged little substantive progress, particularly on framework conditions for innovation-related cooperation. Against that backdrop, the Commission decided for the first time to apply Article 22(6) to China, narrowing Chinese access to Horizon Europe innovation actions. The Commission’s webpage on international cooperation with China was updated to state that, pending agreement on the roadmap, Chinese entities may in most cases only participate as self-funded associated partners.

The authors document how this logic reaches down to individual funding calls. In a 2026 health-cluster call on anticipating the impacts of climate change on health, with an indicative budget of 55 million euros, the work programme text explicitly states that legal entities established in China are not eligible to participate in the relevant research and innovation actions or innovation actions, citing stalled joint roadmap negotiations and concerns over undesired intellectual-property transfers linked to China’s civil-military fusion policies and data-security legislation. A civil-security cluster call on stress-testing critical infrastructure, with an indicative budget of 10 million euros, goes further, limiting participation to entities from member states and associated countries and barring any entity directly or indirectly controlled by non-eligible countries, including China—rules the authors say preclude EU-China joint work on disaster resilience.

By late 2025, leaked drafts of the Horizon Europe 2026-2027 work programme, reported by Science|Business, The Register and Table.Media, suggested that Chinese entities might be barred from a significant number of calls in key clusters such as health, civil security and digital technologies, citing trade-secret protection and the unwanted transfer of dual-use knowledge. The drafts reportedly also single out the seven universities known as the “Seven Sons of National Defence”, linked to China’s defence industrial complex and overseen by the Ministry of Industry and Information Technology, as ineligible across the programme on the grounds that Horizon Europe has a strictly civilian-only mandate. The authors caution that the analysis is confined to publicly accessible texts and makes no causal claims about implementation behaviour or research-output effects.

Crucially, the study finds a dual trajectory rather than a rupture. Established cooperation architectures persist: the High-Level Dialogue on Research and Innovation in January 2021 reaffirmed the importance of research and innovation in the broader relationship, and in November 2024 the sixteenth EU-China Joint Steering Committee meeting reiterated commitments to the co-funded flagships and willingness eventually to finalise the joint roadmap. Yet these channels are increasingly overlaid with security-oriented constraints that tighten eligibility and risk parameters. Cooperation has not been terminated, but is increasingly premised on selectivity and security-conditional openness.

China’s response, the authors show, has been primarily discursive. Chinese officials and state media have developed counter-narratives that reframe de-risking as “decoupling in disguise”, portray it as politically motivated containment, warn that it carries high risks for Europe’s own competitiveness, and argue that an excessive focus on risk means missing out on opportunities. A widely circulated May 2024 Xinhua commentary under the pen name “Xinping” warned that an obsession with de-risking could undermine industrial competitiveness, fragment the single market and erode strategic autonomy, drawing on European business voices to argue that over-securitised risk discourse distorts cost-benefit calculations. Statements by Chinese diplomats, including Qin Gang’s warning that de-risking may become “de-Sinicisation in the name of de-risking” and Wang Yi’s insistence that for Europe China is an opportunity rather than a risk, directly contest the EU’s framing.

Yet the picture is not symmetrical, and the authors are careful on this point. Domestically, Beijing has simultaneously tightened its own data and research-security frameworks, including the Data Security Law and stricter regulations on cross-border cooperation involving human genetic resources. The resulting dynamic is described as an asymmetric, interactive process of mutual knowledge securitisation: an externally oriented, rule-based EU regime intersecting with mainly discursive Chinese resistance and selective internal tightening of knowledge controls, with “mutual” denoting interaction rather than institutional equivalence. The authors also note that knowledge securitisation is uneven within the EU itself, since member states differ markedly in their willingness and capacity to implement security reviews, producing a patchwork rather than a uniform regulatory regime. The normative stakes, they conclude, lie in managing real risks—espionage, intellectual-property leakage and the militarisation of research—without normalising exceptional measures, stigmatising partners, hollowing out open science or pushing collaboration networks to relocate away from Europe.

Subject of Research: How the EU's de-risking agenda securitised research cooperation with China between 2021 and 2025

Article Title: EU–China research cooperation under a “de-risking” framework: the securitisation of knowledge in EU science policy (2021–2025)

Article References: Li, Y., & Long, T. (2026). EU–China research cooperation under a “de-risking” framework: the securitisation of knowledge in EU science policy (2021–2025). Asian Review of Political Economy, 5(1), Article 13. https://doi.org/10.1007/s44216-026-00080-0

Image Credits: AI Generated

DOI: 10.1007/s44216-026-00080-0

Keywords: EU-China relations, Horizon Europe, de-risking, research security, knowledge securitisation, science policy, Article 22(6), securitisation theory, science diplomacy, Chinese counter-narratives, research governance, economic security