Research security is widening in scope faster than institutions can build the skills, tools and clarity to manage it. The central paradox facing science today is sharp: how do you protect critical R&D without suffocating the global collaboration it depends on? That was the question at the heart of a recent roundtable in London.
In October, Digital Science and Ulterior hosted an executive roundtable on securing R&D intelligence for UK and European funders and government institutions. The session brought seventeen guests from across government, funding bodies and research organisations together with the Digital Science team, including Dr Leslie McIntosh (VP, Research Integrity and Security) and Marko Zovko (Global Business Development Director, Government, Funder and Non-profit).
The discussion focused on three core areas: the current landscape, practical challenges and best practices, and openness versus security. What emerged was a portrait of a research sector grappling with fragmented data, shifting definitions of dual use, and an urgent need for network-level clarity.
The conversation ran without attribution, so what follows reflects what we heard around the tables, not the views of any individual or organisation.
Research security now reaches further
Dr. Leslie McIntosh set the scene by showing why research integrity is fundamental to research security, and how patterns in affiliations and collaboration networks can signal risk.
Dual use came up repeatedly. Some guests noted that certain funding is increasingly framed around dual-use potential, and that the concept is broadening into areas such as AI. Because the eventual application of a project (drones, for example) is often unclear, researchers may end up working on dual-use projects without realising it. Guests also described a tension between government pressure to produce more defence-relevant research and the need to secure it.
Sovereignty pushed its way up the agenda, with guests stretching it well beyond geopolitics to cover data, IP and supply chains. Some guests raised concerns about IP, data assets and personal data coming under foreign or corporate control, and about where industry chooses to locate. Others described diversifying data services, including away from the largest commercial cloud providers, as part of staying safe, and saw “middle power” countries as a growth area for trusted collaboration.
Horizon scanning drew broad interest, along with a sobering framing. By the time a signal is visible, the next development has already begun. The “next best thing” is watched and protected, while the “next, next best thing” is unknown and at high risk.
The practical burden falls unevenly
The biggest gap guests named was skills and capacity. Upskilling researchers and institutional staff on research security is costly, and people already have little time for grant paperwork. Due diligence requirements can slow or stop research, with smaller, less-resourced institutions facing a heavy disadvantage.
Sharing is hard, abroad and at home
Due diligence gets harder across borders. UK requirements can be incompatible with overseas partners’ policies and capabilities, which leaves researchers stuck and, some guests felt, places heavy demands on the very countries a grant is meant to support.
The problem is no easier closer to home. Silos make it hard to move data inside government organisations, let alone across agencies and research bodies.
The blind spot after the award
One of the biggest risks comes after the grant is awarded: guests described vetting of downstream collaborations, such as grant-funded start-ups and spin-offs, as extremely rare, even though they saw it as a major security concern for processes, personnel and IP.
Incentives play a part
The publish-or-perish mentality persists, with few other ways to build credibility, though research blogs and alternative metrics were mentioned favourably. Research departments are increasingly expensive to run, and some guests felt that industry and defence bodies are using PhD students as a lower-cost research resource. Others said that unclear IP and data ownership holds back innovation before IP is even created.
From project-level checks to network-level questions
The European Commission’s principle of “as open as possible, as closed as necessary” was a natural reference point, not least because guests warned of a chilling effect on legitimate collaboration. Restrictions on foreign students, for example, aim to reduce risk but can hit institutions that depend on international students financially and for their research. We argued that making the principle real takes more than a compliance framework. Project-level questions, such as whether a grant is compliant, a partner is on a restricted list, or a researcher has disclosed all affiliations, can be answered, but they are not sufficient.
The harder questions sit at network level: where capability is concentrated across a portfolio, which collaborations are load-bearing and which are peripheral, and where dependency is forming without anyone intending it. Marko Zovko set out what answering them requires: persistent identifiers used consistently across borders and agencies, linked data across funding, outputs and collaboration, and analytics that can surface what manual review cannot see at scale.
The regulatory picture adds urgency. Research security requirements under FP10 (the EU’s next framework programme), the EU Research Security Centre, a European due diligence platform and UKRI’s Trusted Research and Innovation principles are all moving forward at the same time. Applying “as open as possible” is what strategic governance looks like for a research system that has to stay internationally connected.
Key takeaways and next steps
Guests valued the mix of perspectives around the tables and the chance to meet peers from other organisations. One described the evening as “a good opportunity to engage with other professionals and explore the topic from different angles”. Several practical steps emerged from the discussion:
- Ask network-level questions alongside project-level compliance checks.
- Use persistent identifiers consistently across borders and agencies, and link data across funding, outputs and collaboration.
- Invest in awareness training for researchers and institutional staff, particularly as dual use spreads into areas such as AI.
- Support smaller institutions that lack the tools and capacity of larger universities.
- Extend due diligence downstream to grant-funded start-ups and spin-offs.
A recurring reminder was that research security is about the motivations of the people involved, not only the research and its outputs. That means knowing the people behind the research as well as the research itself, and making that possible without making collaboration harder.
Our thanks go to Ulterior for partnering with us, and to every guest who joined us in London.
Balancing international collaboration with national security starts with connected data. Learn how Dimensions Research Security supports funders and institutions, or arrange a demo with one of our experts.
