The 2026 Annual Report: Which Universities Are Most Linked to the Arms Trade?
by Vicky Stubbs
Nov. 16, 2026
University relationships with the arms industry extend beyond investments. They include research collaborations, sponsored studentships, departmental partnerships and recruitment activity that helps arms companies reach students.
At Demilitarise Education, we investigate and make public these relationships through our Universities and Arms Database. Our 2026 Annual Report brings that evidence together to examine the scale of university involvement, the limits of disclosure and the changes needed to make institutions accountable.
The rankings provide a starting point. Understanding how we gathered the evidence - and what universities did not disclose - is essential to using them responsibly.
Read our Universities and Arms Database Annual Report 2026.
What does the Annual Report cover?
The report examines documented university relationships with arms companies and military bodies, using an eight-year reporting window covering partnership start dates from 1 January 2018 to 31 December 2025.
That window matters. The figures describe evidence recorded across a defined period; they are not a statement of what each university currently owns or receives in a single year.
Our methodology includes qualifying partnerships in aggregate totals according to their start date. A partnership remains within that window regardless of its recorded end date. For investments, the dates generally refer to financial reporting periods rather than the precise dates on which shares were bought or sold.
Readers should therefore check the underlying record before describing a relationship as current.
What kinds of relationships do we record?
The report brings the evidence into two main categories: direct investments and academic partnerships. Research is included within academic partnerships, alongside other relationships involving teaching, recruitment and institutional support.
Direct investments
We record direct holdings of stocks, shares or equity in named arms companies.
“Direct” describes the holding, rather than who manages it. A named company holding can be managed externally and still fall within this category.
Pooled funds present a different research problem. Their underlying holdings can change, and university disclosures may not identify the individual companies or the university’s attributable exposure. Our direct-investment figures should therefore not be read as a complete account of possible indirect exposure.
Where no direct investment is recorded, further questions about pooled funds may still be necessary.
Academic partnerships, including research
Academic partnerships encompass research collaborations and other relationships through which arms companies and military institutions engage with universities.
These can include:
- Research projects involving a named arms company or military organisation
- Sponsored PhDs, studentships, scholarships and professorships
- Apprenticeships, graduate schemes and recruitment
- Careers-fair appearances and job advertisements
- Departmental relationships, consultancy and event sponsorship
Some involve disclosed financial values. Others provide access, institutional credibility or recruitment support without a recorded monetary value.
The report explains that we are bringing research and other academic partnerships into one overarching category because university disclosures frequently overlap. A sponsored PhD, for example, can involve both research and student funding.
Research figures must therefore not be added to an academic-partnership total that already includes them.
Which universities appear highest in the report?
The report’s overall ranking places the following institutions in its first five positions:
| Position in the report | University |
|---|---|
| 1 | University of Sheffield |
| 2 | University of Bristol |
| 3 | University of Manchester |
| 4 | University of Birmingham |
| 5 | Imperial College London |
The ranking is based on recorded monetary values within the report’s scope. It is not a ranking of current shareholdings alone, nor a complete measure of every university’s involvement in the arms trade.
An institution with substantial disclosed research relationships may appear above one that has provided less information. A recruitment relationship can also matter ethically while contributing little or nothing to a financial total.
The useful next step is to examine the records behind a university’s position: which organisations are involved, what kind of relationship exists, which dates apply and what the evidence actually establishes.
How did we gather the evidence?
Freedom of Information requests are central to our research. We use university responses alongside public sources, including investment disclosures, financial reports and research-project information available through UK Research and Innovation’s (UKRI) Gateway to Research.
For this reporting cycle, research collection began in April 2025. We tracked requests and responses, recording disclosure levels, delays, clarification requests and exemptions.
Our broader research criteria draw on the Top 100 SIPRI and Defense News lists of leading arms-producing companies. To make requests more manageable, the 2025 process defined an “arms company” using those source lists, but then used a narrower list of 20 companies for investments, with the Ministry of Defence, Defence Science and Technology Laboratory and Atomic Weapons Establishment added for academic partnerships.
This scope has limits. It does not cover every company connected to military production, and universities did not always interpret requests consistently.
Some searched broader company lists; others searched the narrower selection. The periods covered by responses also varied if time/cost limitations applied to the full timeframe. These differences affect how confidently institutions can be compared.
Keeping an audit trail allows readers to examine where information came from and enables us to revisit records when universities provide corrections or further disclosure.
A named partner is evidence of a relationship — not proof of every possible end use
Research can have both civilian and military applications. However, a broad subject label such as “dual use” does not establish an arms-industry relationship.
Our methodology requires a demonstrable link: an arms company or military organisation must be explicitly named as a funder, partner or collaborator. A project funded solely by a research council does not qualify simply because it sits within a category such as aerospace or defence.
The distinction between partner and funder also matters. An arms company may participate in a publicly funded project without providing its entire budget. A project’s total recorded value should not automatically be described as money paid by that company to the university.
Likewise, a documented partnership does not establish precisely how every research output was used. A small development in a specialist field may contribute to many later applications, and the university itself may not know all of those eventual uses.
We should distinguish information that is held but not disclosed from uses that are unknown or cannot yet be traced. Our scrutiny should focus on the partners involved, the agreed research aims, the potential applications considered and the ethical review undertaken.
We will explore the complexity of research end uses in more detail in a future blog post.
Why missing information must remain visible
A low recorded value can have several explanations. It might reflect limited qualifying relationships, undisclosed values or incomplete information.
Our report distinguishes between several situations:
| Recorded situation | What it means |
|---|---|
| No qualifying partnerships confirmed | The evidence supports an absence within the scope and period examined. |
| A relationship with a £0 monetary value | A link exists, but it may involve an activity such as a free recruitment advertisement. |
| A relationship whose value was withheld | The connection is evidenced, but its financial scale is not disclosed. |
| Incomplete or missing disclosure | The available evidence does not support a confident conclusion about the full extent of relationships. |
These situations should never be collapsed into a single claim that a university has “no arms links”.
Equally, non-disclosure does not establish the value or nature of an undisclosed relationship. It establishes a gap in public accountability that requires further investigation.
What did the research process teach us?
The report’s lessons and limitations are part of the findings. They explain why obtaining a clear picture of university arms relationships remains difficult.
Records are often scattered across departments
Research, studentships and other partnerships are not always recorded in one central system. Information may sit within individual schools, departments or research offices.
This can make institution-wide searches difficult and contribute to incomplete responses. For our next report, we will ask universities for data held centrally to make requests more manageable.
Centrally held records may still omit relationships recorded only within departments, so we will make that limitation clear. Universities should improve their institution-wide oversight of these relationships.
Why reporting periods vary and what that means for comparison
We ask every university for the same standard period of eight years. Some universities refuse on cost grounds under Section 12 of the Freedom of Information Act, which lets them decline a request if finding the information would exceed a set time and cost limit.
When that happens, we narrow the request, usually by shortening the timeframe, and resubmit. Our priority is to get as much information into the public domain as possible. A shorter period with some disclosure is more useful than a full period with none.
Universities all report on the same financial year, from 1 August to 31 July, so the years themselves line up. What varies is how many years each university discloses. Universities also keep records for different lengths of time under their own retention policies, so older information may no longer exist.
This limits year-on-year and cross-institution comparison in a few ways:
- Totals reflect what was disclosed, not necessarily the full picture. We send similar requests every year so that, over time, we build up eight years of data for each university. Where a university has used exemptions leading to less disclosure, its total will be lower. This is why our report compares universities over the full eight-year period and also over the 2023/24 and 2024/25 financial years alone, where coverage is more complete.
- A shorter record does not mean fewer ties. It may only reflect a narrowed request.
- Trends are most reliable within a single institution, and only where it disclosed the same years across our requests.
We rank universities by the total monetary value of their disclosed ties. Each ranking should be understood alongside the barriers to disclosure.
A university’s position reflects what it disclosed to us, and a lower position may partly reflect less transparency rather than fewer ties.
Database entries are not always individual partnerships
One entry can contain several activities, such as a university’s grouped disclosure of job advertisements.
The 2026 report identified better handling of grouped records as an area for improvement. This is why partnerships now contain a quantity field, in order to count individual relationships or recruitment events under one single database entry.
Gaps in disclosure make change over time harder to track
The report looks at new partnerships started in the 2023/24 and 2024/25 academic/financial years. Even in a narrowed two-year request, many universities did not disclose everything we asked for.
A change in recorded activity could reflect funding decisions, reporting differences or newly available evidence. A fall may mean fewer ties, or less disclosure; a rise may mean new partnerships, or more openness.
We still identify trends and give context for why more projects may start in one year than another. Where FOI disclosure falls short, we supplement responses with UKRI’s publicly available research funding data.
We do not present a change in one university’s figures as proof on its own that it is militarising or demilitarising. Universities can help close these gaps by disclosing the information they hold.
How students and staff can use the findings
Our aim is to turn evidence into specific, achievable demands.
Start with your university’s records. Identify the company or military body, the relationship type, the relevant department and the evidence date. Check whether the source establishes a current relationship or a historical one.
Then use that evidence to ask focused questions:
- Which recorded relationships remain active?
- Who approves them, and what ethical criteria apply?
- Are students and researchers told who funds or participates in their work?
- Does ethical review consider the potential uses of research outputs and are students and researchers made aware of these potential implications before agreeing to participate?
- Which employers receive access to careers fairs, job boards and departmental events?
- Will the university publish its investments and partnerships routinely?
Bring the findings to student and staff unions, departmental meetings and university governance bodies. Where disclosure is incomplete, seek a clear response and a timetable for publication.
Our Demilitarise Education Treaty provides a framework for turning these demands into institutional policy. The purpose is to secure lasting changes in how universities invest, conduct research and facilitate recruitment.
Explore the evidence behind your university
Our Annual Report makes documented relationships easier to examine and challenge. Its value depends on reading the figures alongside their sources, definitions and limitations.
We want universities to account publicly for the partnerships they choose and the interests those relationships serve. Transparency is an essential step towards ending university complicity in the arms trade.
Read our Universities and Arms Database Annual Report 2026.
Explore the Universities and Arms Database, examine your university’s records and use the evidence to campaign for education that serves peace.