Welcome to the Public Accounts Committee on Thursday 11 June 2026—the year is flying by. The Department for Science, Innovation and Technology, DSIT, is responsible for positioning the UK at the forefront of global scientific and technological advancements, and it sponsors UK Research and Innovation. UKRI is a non-departmental public body sponsored by DSIT, and it is responsible for almost all the UK’s research infrastructure, the facilities, resources and services that researchers use to carry out their work. In ’24-25, UKRI spent £1.2 billion on building, maintaining and upgrading research infrastructure, representing 10% of total DSIT spending on research and development, which was £12.6 billion. The DSIT R&D budget for ’25-26 is forecast to grow to £13.9 billion. UKRI also established the infrastructure fund to provide a more structured and strategic approach to identify, prioritise and select major research infrastructure projects for investment. UKRI has approved 30 projects at an expected cost of £2.4 billion. Last year, however, our Committee reported insufficient clarity on how UKRI invests its money. We called for the establishment of clearly defined objectives from the organisation against which to judge its performance and to hold it accountable. Today, therefore, we will examine whether DSIT and UKRI ensure that spending on research infrastructure is aligned with national priorities and is managed appropriately. We will also explore whether DSIT and UKRI have effective processes to support funding decisions or could do more to realise the potential economic benefits of public investment in research infrastructure. To help us, we have an experienced group of people. I welcome Emran Mian, the DSIT permanent secretary—will you introduce yourself first and say when you were appointed?

My name is Emran Mian. I am permanent secretary for the Department. I was appointed in July 2025.

Good morning, everyone. My name is Sir Ian Chapman. I am the chief exec of UK Research and Innovation. I was appointed in August 2025.

Hello. I am Michele Dougherty. I am chief executive of STFC, but I am also UKRI champion for both infrastructure and space. I was appointed in January ’25.

I am Alexandra Jones. I am director-general for growth, science and international. I was appointed in June 2023.

You are all very welcome. Thank you for coming today. I think you have all been warned—I will address this to Sir Ian to begin with—but the Science and Technology Facilities Council is concerned about its budget, because it has some pretty high-profile projects, such as the large hadron collider at CERN, the high-luminosity LHC to begin operation in the 2030s, the UK leadership of ATLAS and CMS, and a number of other high-profile questions. We received a letter from Dr Simon Williams, a postdoctoral research associate of the Institute for Particle Physics Phenomenology at the University of Durham. He asked us to ask two questions. First, Sir Ian, given the UK’s continuing participation in those important international programmes I have just mentioned, what value-for-money assessment has been made of cutting the domestic research workforce that turns those commitments into scientific, technological and economic return?

I will give some context first. The STFC budget for ’25-26 is £835 million; by the end of the spending review period, it will be £842 million. That is not much of an increase, so it is an essentially flat budget over the four-year spending review period. The misnomer is that the budget is being cut; the budget is flat. Conversely, our forecasts of the costs against our existing portfolio of activities are that the costs will exceed the budget; the budget has not been cut, but we are having to rationalise our portfolio to make sure that our spending meets the budget. That is just a rational and reasonable management thing to have to do. We are going through a process on that and we have not yet determined exactly which parts of our portfolio may need to change. We have been interacting with the community over the past six months. We identified this problem and started talking to the community in October ’25. We issued letters to all grant holders to look at different funding scenarios in January. Recently, we have received all that feedback. Now we are doing detailed impact assessments, as your correspondent rightly asks for, and we are in the final throes of that. We had a council meeting of the STFC last week; we have further science board meetings to come and the UKRI board is meeting at the start of July to look at the different options for how to manage the cost pressures. After that, we will feed up to Ministers and the Department before we can communicate onwards to the community about exactly what we intend to do. We have, however, taken those six to nine months to do detailed assessment, talking to the community. We, with Michele, did so many town halls and events to talk to the community and to consult on their priorities and what the most important things in our portfolio are. We have the detailed impact assessments now, underpinning these decisions. We have not rushed into anything here. We are taking our time to make the right decisions.

Nevertheless, you have given the figures, and £835 million to £842 million is virtually a flat budget; with inflation, that amounts to a cut. The worry is that we will not be able to participate fully in projects like CERN, because we will not be able to afford the number of researchers needed to play Britain’s part hitherto in CERN. Therefore, the second question to ask is, what is the exact PPAN-wide change in the funded postdoctoral position before and after the reprioritisation?

We have committed to keeping the FTE postdoc numbers at 2025-26 levels. It will be broadly consistent with the same number of postdocs that we have now. We have committed to do that as part of the work we are doing.

Could you let us have a note on the amount of postdoctoral research posts there will be? What I am getting from Sir Ian, and I might be wrong, is that this budget has not yet been finalised.

That is correct.

When you finalise it, could you please let us have a full note? This is the one research council that we seem to have had a lot of correspondence on. Clearly, those people are very unhappy about what is going on.

Yes.

Moving on to the hearing proper, in its opening remarks the NAO says that the UKRI does not have an overarching framework to monitor and evaluate the performance of research infrastructure and that a more consistent approach could generate additional support decisions. Considering that you are a scientific body that is going to invest a great deal of money in digital analysis, I find it very strange that you cannot do a concrete analysis of the business cases and their effect on research and growth in this country. I understand and sympathise with you because you have very difficult decisions to make. However, unless you measure those things in some quantifiable way—as we would like to see—we do not know how you make those decisions.

I will start at the top level. DSIT has a framework document with us where they set us objectives. Those five objectives were published last year, and they are pretty clear about what is wanted from the organisation. That then flows down into a strategy for UKRI for the next five years that is hung around those five objectives—as you might imagine—and which we will publish later this year. Each of the five high-level objectives will have a couple of key performance indicators against which we expect to be held to account over the next five years. Below that, for each of those five objectives, we then have a series of much more granular details on what we are aiming to achieve. Later this year we will publish a delivery plan for the first year underpinned by another 50 metrics. Of those there are about 10 that the Department holds me to account on. Those are grand, high-level macroscopic measures. There are then about 50 metrics that I hold my team to account for and for which I can identify individuals to be held to account against. Concerning infrastructure, some of those measures are around the performance of our infrastructure portfolio. The performance of the infrastructure is no different to the performance of our skills, research or commercialisation portfolios. All of them tie back to the mission and purpose of the organisation, which is to advance knowledge: how do we discover new things that are driven by curiosity and that allow us to push the frontiers of human knowledge and improve people’s lives; how do we make social impact on the public—who pay for us—and how do we improve their lives and drive growth? That is something that I think has been less overt in UKRI’s mission in the past, but we see it as an integral leg of what we are trying to achieve. Those three legs are all equally necessary and we need to look for the symbiosis between them and how they are self-reinforcing. You would not just do discovery research without pull-through into commercial benefit. You cannot try to improve people’s lives without doing discovery research. You have to do all three. To your question on the infrastructure portfolio, that is how we hold our investments to account. Are they making social impact? Are they delivering commercial benefit? Are they advancing curiosity-driven research that discovers new science? Those are our three macroscopic measures. Then at a more granular level you can look at publications and citations. There are various metrics you can use for scientific impact. You can make measurements about social impact, the number of jobs supported and communities engaged with. You can make measurements about place impact. You can look at the pounds, the number of companies spun out or supported by an infrastructure or the number of people that they employ. We measure all those things. We have a very detailed measurement and evaluation framework. The right criticism is that, while we do that, we have not done it in a completely consistent way across our portfolio, and we are moving toward that in the future—a completely consistent monitoring and evaluation process.

I will give an example from this week. As part of London Tech Week, we have published our AI hardware plan. Within that, we talked about a piece of research infrastructure that we are funded to deliver in this spending review period, which is AI research resource—over £1 billion of investment into high-performance computing. In a previous time, we would have been relatively agnostic as to what chips go into this system, and we would simply go to the market and buy what seemed to be the highest-performing chips at the time. As everyone knows, at that time, that would probably end up being chips from one or two American companies. But given we have a cohort of UK companies that are developing the next generation of chips—we have supported some of them through early UKRI projects and some through ARIA projects—we very explicitly said that we want the system we buy to be one that accommodates different kinds of chips and is not a single kind of chip. That creates the opportunity for us to build this system including chips from British companies. That provides those British companies with the next step in their growth journeys, so they go from being small R&D operations to having their first scale customer. Once they have demonstrated value there, we would hope that they are able to scale and sell to other big customers as well. I hope that is a helpful example of the shift we are making that Ian was describing, where we are linking our research spend much more explicitly to our industrial strategy.

I hugely welcome that answer, because I think that anything that strengthens UK infrastructure resilience in this uncertain world, so that we have a capacity to produce these cutting-edge chips—chips are the basis of virtually everything we do—is hugely important. A very welcome answer indeed.

I have to say that I read this report with interest. I find myself confused by the all the disparate funding streams: the majority go through UKRI, but there are other bits and pieces and it all seems to be quite disjointed. The objective of my question is to make sure that we understand that you know that you are spending the right amount of money on infrastructure. To that point, the NAO came up with this in the intended objectives for the session: “UKRI does not have an overarching framework to monitoring and evaluating the performance of research infrastructure.” If we look at paragraph 3.6 on page 37, it seems to me that you rely quite heavily on the universities for infrastructure. I did a bit of work on the universities and, if you are relying on them, I think that you potentially have a problem. When I looked at the universities, I found that many of them are actually suffering very large deficits, which you probably know: Cambridge, £53 million; Coventry, £60 million; Brunel, £56 million; Essex, £22 million; Derby, £23 million and Cardiff, £31 million. I am also concerned that they seem to be spending hugely increased percentages on non-research-related jobs; they have been hiring or bringing in people to do roles such as senior continuous improvement managers, associated director of belonging and engagement, events and experiential learning officer and immigration compliance associate, and the research element versus the administrative element has dropped. Equally, we read here that £5.6 billion is required to restore university-owned research infrastructure to a fully operational condition. Even though I think what you are doing slightly smacks of central planning, which I never like, do you think that you are investing your money wisely in the correct infrastructure, given that there appears to be a weak link with the universities? Are you chucking good money after bad?

Who are you addressing that to, Rupert?

Probably to Sir Ian and Emran.

There are many facets to that question. Can I start with how we distribute our money? I sympathise with your viewpoint that it is quite hard to interrogate how the money gets spent. To that end, from 1 April we are completely changing our financial ledger system. Previously, we had a large number of relatively small schemes, and we are moving that through to aligning our ledgers with our mission so that you can see the connection from strategy to spending—it is really very overt. The first bucket of money is around curiosity-driven research and discovery. The second bucket of money is aligning with societal and Government priorities—predominantly, how we deliver against the industrial strategy. The third bucket of money is helping companies to start, to scale and, ultimately, to stay and prosper in the UK. Our schemes will align with those three high-level missions, which the organisation has been set by the Government. I hope that in the future it will be much clearer to see how our money aligns with those outcomes. We are trying to drive spending to align with outcomes, which I think is the thesis of your question. On our exposure with our partners, of course UKRI itself performs research—we have about 6,000 researchers, which is sort of the scale of the University of Oxford or the University of Cambridge—and we operate a large number of research infrastructures ourselves. But the predominant amount of our funding—more than 80%—goes into universities, and we work in concerted partnership with them. I don’t think it is my job—and I don’t think you want it to be my job—to micromanage how they achieve their outcomes or to look at exactly their staffing mix, how they are setting up their employment contracts and who they are employing. I should be holding them to account against outcomes. I should be saying very clearly, “This is what I want you to achieve with that investment of public money,” and then holding them to account against those outcomes; how they achieve them is their business.

But you are relying heavily on the universities for the infrastructure investment, unless I have misunderstood it.

Totally correct.

And when you do not have a fully integrated portfolio review—I read that you do not—how can you tell that the money going into universities is not going to fund this administrative burden that they are taking on? With Cambridge, we read about the influence of China—funding streams coming from China. If universities are short of money, they become open to arguably more malign influences, such as funding from overseas countries that may not have the interests of the British people at the top of the agenda. I think it is a valid question for the Committee to ask. I always think, Sir Ian, that when you are dishing out large amounts of money, as we have known from history, it opens up the risk of fraud, deceit and all sorts of human weaknesses, which basically go with pots of money being dished out. You only have to look at the money being transferred from northern Italy to southern Italy and other issues like that.

I wonder if I might say a couple of broader things about university finances, and then I will ask Ian to say more about how UKRI monitors its spend into universities. The first thing I would say is that the balance sheets of some of the universities you listed there are very strong. The deficits that they may have in any one year might be a deliberate decision because they are investing particularly hard in that year, but they are doing so on the basis of a balance sheet that means that that is a perfectly legitimate decision for the university to be making.

From what I can see, it has been going on for some time.

We should not be looking to control that, because they are independent institutions. Equally, for some of the universities you name, the deficits probably are worrying, and they may not have the balance sheet position that means that they can run a deficit like that year after year. The second thing I would say is that we are looking very hard at university financing and sustainability. Ministers made the decision as part of the spending review to protect and, indeed, to continue to increase the proportion of research funding that goes to universities as a block grant to ensure there is that stable footing for universities. Tuition fees, after a long period of being flat for domestic students, are going to increase again in line with inflation. The other thing that universities would say is that this financial position is driving some different business decisions by them. We have seen some key decisions about consolidation in the last few weeks; King’s College London and Cranfield University have announced that they will be merging some of their operations, and we have seen other instances of that. I do not think that it is for us as a Department to adjudicate on those. They are independent institutions, and it is for them to make those decisions, but those decisions, especially that one, feel like a good way of protecting the research strands of both those universities.

As you know, with merger comes dislocation—I do not need to tell you that. There is £5.6 billion needed; clearly there is an issue of under-investment, which is affecting research. Is that not what the NAO is saying in paragraph 3.6?

Do you mind if I come back on a few things there? To start with, we should remind ourselves that our universities are a huge national asset. The fact that we have four universities in the top 10 in the world is a massive asset for this country, and we should exploit it more. It is not just a few; we have 17 in the top 100, so we have both depth and breadth. I will pick up on your comment about our approach to portfolio management. We really do oversee all our infrastructure as a portfolio, be that physical infrastructure, digital infrastructure or maintenance of infrastructure. They are seen as a portfolio of investments, which is the right thing to do. That allows us to change spend profiles on individual projects so that we can meet an overall spend profile. We do that actively and manage it actively as a portfolio. To your point about the condition of the estate—

I think it is a question rather than a point.

Absolutely. The Report rightly acknowledges that to maintain the existing estate we have a deficit. What it does not assess is whether we want to maintain that existing estate. That is all the infrastructure that exists at the moment, but we should constantly be making active priority decisions and saying, “Do we want to maintain that piece of infrastructure that we invested in 20 years ago, or do we want to now offboard it?” We should be constantly making those decisions. That is not addressed in the Report; it just mentions maintaining the existing portfolio. The portfolio of things that we want to maintain is smaller than that number, and therefore we think it is manageable within our settlements. You asked about China and fraud, and I will address those two things. When we make investments into universities, we have a very clear policy on what we call trusted research and innovation, which outlines the expectations of our investment and how that is protected, how it is used in the nation’s interests and the partners with whom we can work on different subjects. There is a very clear policy on that, which all our investees, universities, centres, catapults or companies have to abide by. We take fraud very seriously. We are spending public money, and any of the endeavours and ventures are very early stage and risky. To give you comfort, for the 2024-25 financial year, our fraud rate was 0.01%, which is very low. Error was bigger than fraud, but even our error rate was 0.1%, most of which we recovered. We do a very good job of managing fraud over a big portfolio.

On that basis, you are obviously happy that you are directing your investment into infrastructure correctly.

Yes. We have an exceptional portfolio.

May I come in on this? On page 37, paragraph 3.6, the NAO says that “the Office for Students estimates that 45% of providers are on course to operate with a deficit”. I hear what you say about the particular reasons why somebody is running at a deficit, but you also highlighted the Cranfield-King’s merger. I suspect that is likely to go on, and there will be more consolidation in the universities sector, particularly if news headlines are right that the number of students is dropping off and people are choosing not to go to university. Sir Ian, are you absolutely sure that you are investing in universities that will still be there in the timescale that these research projects are supposed to produce results? It would be folly to invest money in them and then they cease to exist. Another one of you may want to answer that question.

In terms of the way that UKRI assesses which universities to invest in, it looks very carefully at the track record and the strength of the research, and we have funding streams, like quality-related research, that then double down on that. Where a university is good at something, we give them more money, because they win it and because they’re good at it. There is generally a track record and a reason why it will attract investment and talent, and a reason to go there. Then we have our way of making sure that that investment and infrastructure are sustained over time. Edinburgh is one of the universities that we are investing in, as is Bristol. Those are universities that have long track records and that have very strong research in particular areas, where the infrastructure is being invested. As DSIT we keep an eye on the overall university portfolio, what is happening, and how that works. Generally, however, the ones that get the infrastructure investments are already strong in those particular areas, and we have a plan for sustaining that infrastructure over time, with a combination of maintenance and then enhancement. In a world in which universities are clearly exposed to a number of forces—it is very challenging for them—the ones that we are investing in are also making wider management decisions about how they make themselves sustainable, and about how they make sure that the subjects they are investing in and the infrastructure that they are investing in are sustainable. Most of those will be bidding for this infrastructure on the basis that it is sustainable over the medium to long term. So, they generally have the establishment of the quality; they have the track record. We continue to invest in the research and the infrastructure, because of that track record. And then those institutions take action themselves. You can see that in the mergers, as well.

Okay. That is very helpful. My able deputy, Clive Betts, wished to come in on universities.

Yes, on the distribution of money. It seems that what you are saying, Alexandra, is something that worries many of us about the regional distribution of funding for research across the UK. What you are basically saying is that some universities have had money over the years and therefore they have got very good research organisations. Therefore, they are the ones you give more money to, which means that they become even better at research and get more money. My guess is that the majority of the funding that you give goes to the golden triangle—Oxford, Cambridge and London—and excellent universities in the north get very little. Is that fair?

I don’t think that is fair, actually, and I think it is changing over time. In a moment, I will bring in Ian with some of the figures. What we have seen is that investment in capability across the UK over a long term, through a series of programmes like strength in places, has meant that, through the recent research excellence framework exercise, universities across the UK outside the golden triangle are getting more funding, because the quality of their research has increased. We have also done a lot of work on place within UKRI, so we are seeing an increase over time of investment going outside the golden triangle into other parts of the UK—going to excellence, because there is very high-quality research and innovation around the UK. Then, we also have programmes like the local innovation partnerships fund, which is on the innovation side. That is £500 million. That is just one part of it. A lot of money goes to universities around the UK beyond that money, but with that £500 million we are working in partnership with a number of areas, including Greater Manchester, Glasgow and the west midlands on how we invest in an innovation ecosystem that means they can take advantage of the quality research that we are investing more in, and then kind of spin it out. It is not resolved. I am not suggesting that this is an equal distribution of funding, but we are taking very seriously the issue of: “How do we change that?” And one of the objectives for UKRI is about thinking about its impact on local areas, as well as on the national UK. I do not know if you want to say any more, Ian.

I was going to emphasise that explicit in our five objectives is to think about our impact on place. To show you that UKRI cares about this—we actively care about this—I will go back five years. You used the north of England as an example, so I will build on that example. Five years ago, 12% of our budget went to the north of England and today it is about 20%. We have moved considerably. By the way, that is more than the cross-Government spend benchmark, so we are doing better than the rest of Government on this. We really do care about it and we are actively working with our partners to help support their strategic priorities. The local innovation partnerships fund, which Alex mentioned, is a great example of that. We have identified 17 places that we are supporting through that fund. It is a real co-creation. We are working with devolved regions, or city regions in devolved nations, to identify their real strength—their thing that differentiates them not just in the UK, but in Europe or even in the world. What is their particular area of strength? Then we co-create a programme with them to invest in that strength in that region with them. It is not just us doing it in blissful isolation; we are genuinely working with our regional partners, so that they help to shape that programme and administer it.

I would like to give another example, which is the Infrastructure Advisory Committee that I chair. As part of the assessment criteria that we use, we look not only at the balance of research and infrastructure across subject areas but across location, as well. That is part of the criteria that we use as we look at all the different proposals coming in.

The 20% seems a big figure. Then you look across the north. How many universities in the north share that 20%?

I would have to add that up. I cannot do it quickly enough in my head.

That is the challenge. The northern universities as a whole probably get less than Oxford, Cambridge and London.

Our universities are distributed all over the country.

Perhaps you could give us a breakdown—

I can do that retrospectively, yes.

A breakdown of where the money goes in terms of universities and particular regions and sub-regions, and what that change has been over the last five years.

I am happy to furnish you with that. Like I say, our ambition resides in finding the excellence in places and helping to support their focus.

Do you have a target?

We do not have an arbitrary target. I think the right thing is to look at the outcomes we are trying to achieve as a nation and focus on achieving those outcomes.

The suspicion is that outcomes do not happen unless you have targets to achieve them.

But we have a firm belief that if you do not use all of your muscles, you are never going to achieve above average growth. We have to use all the regions of the country—absolutely.

I will be interested to see the figures that you are going to provide for us on that.

I am happy to do that.

My question is about long-term budgets. Paragraph 2.19 of the NAO Report notes: “DSIT recognises that, in some cases, setting long-term budgets” is the appropriate thing to do. Given that recognition, I am keen to understand why you have made such little use of long-term budgets so far.

I might bring in Alex to speak to the criteria in detail, but we have set out two 10-year budgets so far—one for the National Quantum Computing Centre and one for the Laboratory of Molecular Biology—on the basis that those are facilities and institutions for which there is a very clear long-term case, so they clearly met our criteria. The thing we are weighing it against is how scientific discovery is also changing very fast, so setting in stone a large number of 10-year budgets and using a large proportion of our research budget in 10-year budgets will limit our flexibility to be able to respond to changes in what we ought to be funding to further the objectives of our research funding. Also, the areas we get tied to in our 10-year budgets will mean we will have less flexibility to spend more in other parts of the country. That is the balance that we have been trying to strike.

Before I ask Alexandra to come in, is there a threshold that a project needs to meet, or were the criteria that you referred to widely known and widely understood?

Yes.

It was published in May 2025. It is aimed at specific kinds of activity or institution. That is done for a reason, because it can be the activity is the thing that we need to fund rather than a specific institution. But it has to demonstrate benefit in supporting core national infrastructure; in retaining and attracting talent where the skills development supports growth; and collaboration with other countries, or areas where there is demonstrable need for partnerships with industry that can help drive economic growth. So the criteria are set out. As Emran said, we have funded the Laboratory of Molecular Biology and the National Quantum Computing Centre. We have also had Department for Business and Trade funding driving research and investment in vehicle electrification, and the Aerospace Technology Institute announced they have got 10-year budgets. UKRI sets long-term budgets more generally, anyway, because very few of the infrastructure are limited to the three-year spending reviews, so we can combine that 10-year certainty. Most of those have review points, because that is right—as Emran says, things change. We need to give them the certainty of the funding, but make sure we are clear what we want them to deliver. There are a lot of long-term funded institutions and activities and programmes right across the UKRI programme, and we support them to do that, as does the Treasury, because it is impossible otherwise to have some of the certainty you need to continue.

May I ask UKRI a follow-up question? You have greater flexibility over how you use your budgets compared with other Government Departments. Are there any barriers stopping you managing that money on infrastructure more flexibly?

Building on the previous conversation about 10-year budgets and long-term budgets, I think a more valuable thing in the administration of research and innovation funding is an ability to roll between years. It does not need to be a big ability, but it needs to be some ability, because science never goes as you expect. Unexpected things happen, and if you are constrained to hitting your precise number, it drives perverse behaviour. I would rather have a bit more flex between years than a 10-year budget. I will not name the person, but one of my equivalents in another country says the best thing about science in their country is that they have seven-year budgets, but the worst thing is that they have seven-year budgets. As Emran says, you want to be agile and respond to discovery and what is happening in the market; I think that is absolutely right. I feel the balance is about right and that we have the tools at our disposal. As part of the last spending review settlement, the Department negotiated more flex with the Treasury; they can move 1.5% of the total spend between years. That is about the level we need, to be honest. I feel that we have the freedoms to operate that we require.

Are you therefore saying that you feel that there are not any barriers to managing that flexibly, and that you are comfortable that you have the tools you mentioned?

Yes.

I might just give Ian slightly more operating room on this and say that we were not previously able to give UKRI that flexibility of the 1.5% across years. But understanding the way in which the science budget behaves—particularly the infrastructure budget—and bearing in mind previous NAO Reports about being able to better manage this as a portfolio really helped us to agree with the Treasury that we would provide that flexibility as part of the spending review period. Of course, we are in the first year of that spending review period—it has barely started—so we have just started to operate with that flexibility. We have barely tested whether it will allow us to do everything that we want it to do.

So I should come back in two years’ time and ask you the same question.

Yes. The other thing that needs to go with it is being able to make a commitment into the next spending review period where that is appropriate. We felt it was important to do that on quantum computing, which we did a few weeks ago. We committed to the R&D investment that we are going to make in the next four years—over this spending review period—but we also said that we will set out performance criteria. When a quantum computer meets those, we will buy it in the next spending review period. Explicit Treasury agreement was needed for us to make that commitment, but they were willing to provide that. I was talking about high performance computing earlier, and it goes with the theme of making these advance market commitments in such a way to allow British companies or companies based in the UK to have that pathway to growth.

May I add one thing? I think this is a huge benefit of UKRI’s existence. Because we have brought together what used to be nine separate bodies into one, we can manage as a portfolio. When some things are going over, we can manage that across our portfolio. We can ebb and flow budgets depending on where cost pressures arise. That is a huge benefit of being able to operate a portfolio across the different streams.

May I just ask you, Emran, what co-ordination there is across Government? Clearly you have the biggest research budget, but you do not have the only research budget. You mentioned DBT, and DHSC no doubt has a research budget, so what is the co-ordination across Government and with the private sector? Presumably, certain bits of research are much more attractive to the private sector than others, and Government ought to be thinking about investing where it is not so attractive to the private sector.

This is the weekly joy of Alex’s role, so I will bring her in.

Alexandra, it sounds like you are a specialist.

On working across Government and R&D, yes. We have something called an R&D oversight board, which is co-chaired by me, the Government chief scientific adviser and the Treasury. That is the group that we wanted to bring this together; they bring together all the R&D investors across Government so we can look at this as a portfolio. Over the past few years, we have built the management information you would want to have. It could be improved further, but there is much more information about what we are investing in across Government in terms of R&D and in which areas, and how we can complement one another and make sure that if UKRI is investing in some fantastic research, we are thinking about how that supports and complements where MoD or, for example, DHSC might be investing. That meets on a regular basis, and we have working groups under that, which have been compiling management information to try and help us understand how that portfolio works and how we can tension it. It is also looking at the monitoring and evaluation, what the evidence base is and where it is having the most impact. That is on the Government side of things, and there are various individual boards that happen. Then UKRI has its own programme boards, for example, on the industrial strategy. Each industrial strategy area has a programme board on which members of the relevant Government Department sit, so that UKRI’s work is directly informed. I should say that UKRI attends the R&D oversight board—Ian comes to every meeting on that. For a lot of the work with the private sector, we look at the overall portfolio and think about where the opportunities are. We have an overall target of three to one that we look across, and some of that work will not attract private investment and nor should it, because it is where Government are derisking something and investing where only Government will. Some of it is absolutely about trying to secure that investment and thinking about how we would do that. Where that has worked best is generally at a sector level, working with particular sectors to understand what they are looking for and what their needs are. We have a range of ways to do that with advisory groups. We also have funds. We have the sovereign AI fund. Now we have the NSSIF fund, which is all about investing in secure areas of R&D, so we have ways to work with the businesses where we actually invest in them with a contract or grant. They particularly value, as Emran was saying earlier, contracts of some kind, so we can work with them on a sector basis through those funds that are slightly more general and then through some of the infrastructure that we have. The National Quantum Computing Centre is a great example of where industry has asked, as we move towards commercialising that sector, “Please can we have some of the infrastructure that only the public sector can invest in?” to help them grow in scale. That is a good example of both responding to some of their demands and working with them to make sure that it is used by both researchers and industry.

I have been working in and around the Whitehall research and innovation system for a long time now, and I feel that it is more joined up now than it has ever been. I am very positive about what Alex and the team are leading through DSIT and into the other Departments. As an example of that, not only do we have conversations about what we are all doing, but we also now identify who the most important actor is for any one of the different things. As an example of that, I described before how UKRI has discretised its budget into these three buckets around discovery, applied research and driving companies. That has now been done across all the spending that the Government make in R&D—that same discretisation. It identifies really important things such as the fact that UKRI is really the bastion of the curiosity driven. We make 80% of Government’s investment in discovery research. Although it is half of our budget, it is 80% of the Government’s budget spent in that area. We are so important in that, underpinning the system, and the whole system would atrophy without us playing that role. It highlights what the most important things are for each actor to deliver, which I think is a really good sign.

Let us come to a specific example of what you have just been saying, Sir Ian. This may be a question you want to answer, or maybe Alexandra will. Why did your department ignore multiple reviews warning that the UK’s supercomputing capability was falling behind other countries? Presumably, this is linked to the overspending in 2023-24, when basically your organisation was overspent and a lot of the projects were put on hold. I suspect that this was one of them. Where are we now with all this? Also, I have a nagging feeling in the back of my head that this is an area where perhaps the private sector could play a greater part. These supercomputers get out of date very quickly, but it is the private sector that benefits almost immediately when you put a new one in.

Alex will probably want to comment on this as well, but let me respond from a UKRI perspective. We identified the need in good time. We wrote a business case to make a transition for a new national supercomputing service ahead of the current one ending operations. There just was not enough fiscal headroom at the time that that business case was brought forward to proceed with that project. I am delighted to say that that was part of the spending review—not the previous one, but the one before, when the new Government started. There was a commitment to invest £750 million in a new national supercomputing service. We expect that that will be operational next year. That, however, presents a gap between the current service ending in November this year and the new service at the end of next year. As we could see that that gap could arise, we have been making smaller investments in the near term into a distributed infrastructure of research software platforms, which will allow us to bridge. In compute capacity, we will have more available from November than we do in the current supercomputing service. For the users, we will provide more compute capacity in the near term, and when the new national supercomputer comes online in Edinburgh next year, we will provide considerably more. There should be no gap for the users. We have managed to mitigate that risk.

Where will that new supercomputer that comes online next year be in world terms in terms of capacity? There is a phrase, isn’t there—gigaflops? I think it is one of the measurements used.

Yes—now petaflops.

Megaflops.

I cannot tell you until we procure the nodes. We are about to launch the process to procure the hardware. We have a budget. We will have to see how that procurement goes to tell you exactly what its position will be, but it will be in the top 50 and will certainly give us access to high-quality compute resource again.

We also have the UK’s membership of the European High Performance Computing Joint Undertaking, so we already have access to Europe’s largest supercomputers through that, and researchers are making use of that—Jupiter in Germany, for example. We have also invested in the Met Office supercomputer. We did that a long time ago, back in 2019, so that is something else that we have. Although the Met Office uses that and has been benefiting from that, that is a conversation that we are having across Government.

On my question about the private sector, these things get out of date very quickly but, on the other hand, the private sector gets an immediate benefit from these things. Could you bring more private sector investment in on this particular issue?

That is a good challenge. I refer back to my previous comment about UKRI’s key role being to support discovery research. We are the place for that to happen. There are other data centre builds happening across the country as part of the AI growth zones, of which there is a number now. I expect that almost all those supercomputers will be a split between applications to the private sector and applications to the public sector and to research—perhaps predominantly applications to the private sector. Our role is to support the discovery research, so ours will be focused on that.

It is on discovery research, and on UK start-ups using that supercomputing capacity. We are seeing that happen a lot. For example, with the supercomputer that UKRI funded in Bristol, the Isambard system, we focus really hard on getting start-ups access to that system. Indeed, one of the things that we offer through the sovereign AI fund that Alex mentioned is access to the Isambard computer, because for start-ups in the AI space, compute costs are often by far their largest cost. If we can provide them with access to publicly funded compute, we are sorting that problem for start-ups, and it really accelerates their journey. On the broader picture of the availability of high-performance computing in the UK, Ian is absolutely right: the vast majority of it will be private sector-funded for private sector clients. The challenge there is building it at the right scale. We have lots of data centre capacity in the UK, but often it is at small scale, with smaller facilities than the ones we are seeing in the US or in the middle east, and there are real advantages to those large facilities for large private-sector clients such as the AI firms. We do not currently have those large-scale infrastructure assets in the UK. They will be private sector-funded—they will not be public sector-funded—but we felt that we needed to do more to facilitate the building of those assets, and that is what led to the creation of the AI growth zones programme.

Since you mentioned that these supercomputers and quantum computers are very important to a number of start-up companies, how do you, or whoever runs the existing ARCHER2 quantum computer and the new one, decide who to charge and how much to charge for the use of these things?

On the support for start-up companies, we do not really charge them. They apply for access in the same way that a researcher or university applicant would, and we assess the impact that the use of the computing time will have, be that a research or commercial impact. In the same way that we make an assessment of impact when we make investments through grants, excellence, novelty and impact are the predominant things that we look at. We do the same for compute resource, and we have a panel, so it is not one person but a panel that is discriminating between these different applications. Coming to the NAO Report, that is in a sense why, as our future commitments, we are making a commitment of over £5.5 billion to infrastructure, but over £2 billion of that is going to digital infrastructure. Digital infrastructure allows you to do all sorts of research subjects and help companies as well as researchers, whereas physical infrastructure is usually tied to one purpose. Digital infrastructure is interdisciplinary, with many purposes.

I think the Committee is keen to understand, when you get over-optimistic assumptions and projects that arguably are about to fail, how you actually identify them and deal with that. When I read on page 26, paragraph 2.6: “Of the 24 projects under way in December 2025 and funded by the Infrastructure Fund, 14 (58%) reported that they expected to achieve their agreed outcomes and were not facing any major issues which significantly threaten delivery”, that obviously means that 42% did face issues. So how are you dealing with that?

That is exactly why I said to Sarah earlier that I think the benefit of managing things as a portfolio is that it allows you to take risks. Frankly, we should have some projects that encounter some issues, because we are doing things that have never been done before, and, when you are doing things that have never been done before, you are bound to find speed bumps in the road. The important thing is that we manage those speed bumps and those course corrections, and that we do not overspend on the portfolio. And we have done that: year on year, for the last five years, the infrastructure portfolio has come in on budget because we can manage the ebbs and flows of the challenges that we have when we are doing things for the first time.

On that basis, Sir Ian, how did you get issues such as this one with the Airborne Laboratory, where it looks to me like somebody approved it and then it got junked? How did that happen?

That is a great example, because you have just challenged us to make sure that we are always looking for value for money—and we should always be looking for value for money. We made an investment into environmental science five years ago. The world has moved in five years and we can do many more things using airborne drones and unmanned vehicles, for instance, and the ground-based diagnostics have improved considerably as well. Therefore, the value for money of putting infrastructure on to a plane changed. The costs of operating the plane and infrastructure were going up and the number of users was going down, so the value for money deteriorated over that time. Instead of throwing good money after bad, we curtailed the project. That has released £5 million of unspent money, and we are putting the £30 million of operational costs, which we would have spent operating that plane, back into environmental science.

I accept that, but I ask again: ultimately, somebody made the decision to put that extra money into the Airborne Laboratory; are you telling me that, at the time they made that decision, they did not have the available facts that they subsequently had?

That is correct. The world moves. Things change; technology advances.

So this was basically risk capital that the taxpayer took a punt on and lost; is that what you are saying?

I think, sometimes, we will do that; sometimes we will make investments—

And you think that process is managed fine?

Yes; it is completely robust.

The NAO says, in paragraph 2.9 on page 28, “We have previously observed that costs are more likely to rise if project sponsors have vested interests in ensuring projects are funded (which can lead to over-optimism not being challenged or even to deliberate underestimation).” Is that a concern to you?

It is a concern. That is exactly why we have an independent advisory committee looking at this. And, by the way, over-optimism is a problem for most capital projects across Government; it is not restricted to us. We have two mitigations for that. First, we have independent scrutiny of every assessment that goes through, and we force contingency-management reserves into the budgeting to start with. Secondly—we use this increasingly—we will often do tranched programmes. We will give a tranche of money to do more detailed assessment of the costs, and whole-life costs, of a project. That might take two or three years of seeding money to get to a detailed plan. Rather than relying on a non-detailed plan for cost estimates up front, we will do phasing with a gateway, at which point we say, “Okay, now that we have got full detailed costings, and greater clarity about costs and value for money, do we wish to proceed into the second phase of this?” For sure, that risks some nugatory spend, because we will invest a small amount of money up front and, if the project does not go, aside from training people and developing skills exploring it, you could contend that that is nugatory spend. But it is a small amount, and it gives us a better estimation and a better forecast for managing our forward projects, so I think it is the right thing to do.

So you are reassuring me that you are not working on the basis of what I call the bazaar theory, which is if you buy enough tickets in the tombola, in the end you come out with a cheap bottle of wine. You are telling me that is not the case. Is that right?

Correct.

If I may, the way in which the infrastructure advisory committee works is it consists of a group of independent experts from a whole variety of different areas, who help us make the kind of choices that Ian has just talked about.

I get frightened when I hear independent experts talked about. Innately, it worries me.

What part of it worries you?

Experts who are not necessarily commercialising something always worry me a bit.

We have a lot of commercial expertise on the committee as well—people who have taken ideas that they have generated, followed through with SMEs and built big companies.

Sir Ian, just to put a slight caveat on Rupert’s questions, I would not want you or anybody listening to these proceedings to think that we are against innovation. Innovation involves pushing the boundaries. There will be times when things fail. The only thing I would say on that is that if you are going to fail, fail quickly, learn your lessons and move on.

I am not saying that innovation is a bad thing. I think our job is to try to make sure we are getting value for money for the taxpayer. We all start on that premise. How it is delivered is a different issue. Whether you believe in centrally planned innovation is a different issue. That is something we can all debate at length, but that is not the subject of today.

The question about whether you should have known quicker that you needed the latest technology was valid.

I forgot to say one thing on that. We did make an investment, largely in infrastructure to go on the plane, and 80% of that infrastructure can be reused—either ground-based or on drones. It is not complete nugatory spend as the vast majority of it will be reused, but it was the right thing to do from a value for money and operations perspective to curtail that mechanism for using the infrastructure and use other mechanisms.

That is helpful. I am sorry, I did not give any warning, but it is customary for the Committee to have a break halfway through. It is just after 11 o’clock. If we could be back here fairly soon after 11.05, two of our three Sarahs on the Committee will be asking you questions. Thank you for all your replies so far. We look forward to the second session in a few minutes’ time. Sitting suspended. On resuming—

Sir Ian, when we have decision makers from very different strands of science and tech and you are trying to make decisions about how to allocate budget, how are you able to look at different types of research? How can you evaluate them against each other?

It is a great question. One of the beauties and hardest parts of my job is the disparate nature of the portfolio, which goes from the arts and humanities through to economics and social sciences, and biology through to engineering and physics. It is an incredibly broad portfolio. However, when we make an investment, we do so to advance knowledge, to improve lives or to drive growth, and whether you are making an investment in creative technology or in the life sciences, you are still assessing against the same things. The evaluation metrics are still the same. Are you discovering new research? Are you driving the forefront of our understanding? Are you benefiting the public who have paid for this? How do you impact their lives? Are you creating jobs? What is happening in the place? What does it mean for our place in the world and our international standing? Are you creating companies? Are you helping those companies to scale? What is the timeliness of the investment compared with the sector? What is the additionality of our investment? How do we crowd in money? All those criteria are universal against the disparate nature of our portfolio. I think we can compare apples and oranges, because we use the same monitoring and evaluation criteria and select the same assessment criteria.

If you have competing projects that are scoring highly on all those counts, do you then have to weigh things up against each other when they are in very different spheres? Have you had that experience?

Every time we run an assessment, we have competing projects that are all excellent. One of the hard parts of the job is that we are massively oversubscribed. Every investment we make is massively oversubscribed and we leave high-quality ideas on the table that we cannot fund. That is a great problem to have, actually. The fact that we have a massive oversupply of good ideas into which we could invest is a sign that the sector is in rude health. It is just a difficult job and there is no precision to that.

I wonder whether you can give an example of a recent weighing up between two different projects and explain how you came down on one side rather than the other. Given that it is very difficult, an example would be helpful.

Why don’t I bring it back to the infrastructure fund and the latest wave of infrastructure projects that we have just invested in? Very disparate projects were proposed, and the things we announced that we have invested in are: the national cryogenic facility, which supports development of next-generation quantum technology and superconductors, which have applications in energy transmission and energy projects in general; a platform facility, which is doing imaging across scales, which can be applied in many different sectors, predominantly focused on medical and biological techniques; and two projects looking at clean energy, one in the Orkney islands looking at tidal and renewable energy, and one looking at offshore wind platforms. We made those investments, a range of them in different places, from Scotland down to the south-west. We look through the place lens, the social impact lens and the industrial sector lens—which sectors does the UK have a right to dominate and have a significant market share in—and at things that provide platforms. The imaging across scales, for example, is a platform facility that can be used for many different disciplines, ranging from heritage science right through to medical science. Because we have that portfolio approach, we can tension them, but other projects we could not invest in. We could not pursue a project on a centre for mass spectroscopy. To Rupert’s earlier question, we gave it some seed funding to explore the details, but we could not pursue it. We had a project in particle physics, looking at a next-stage detector at CERN; again, we gave some seed money to evaluate it, but we could not pursue it in this round. We always have an oversubscription and we try to balance those many different, competing demands. Is that a useful example?

Yes, I think so.

Thank you, Sarah. I call the other Sarah, Sarah Hall.

I will ask Michele to answer that, as Daresbury is one of her campuses.

At the moment, what we are doing across STFC is looking at our portfolio. We cannot afford ourselves going to the end of this spending review, so we are looking at all our facilities. We are engaging with our communities to get an understanding of what we might slow down, what we might mothball and what we might need to stop. We have not come to that decision yet, but we are doing it in consultation with our communities, because when we know we cannot afford ourselves, we would not be responsible unless we did that kind of consultation and deep dive on what we do.

No, it is a strategic decision across where we want to be in five to 10 years’ time in the different areas of our expertise. That is why we are consulting not only our scientific communities but our user communities.

RUEDI was one of the projects that was not supported in the December infrastructure round. That was one of the difficult choices that we had to make.

It was very close to the boundary of being chosen, but we were trying to get a balance across not only portfolio but place. That was one of the ones that came just below the line.

No, it is paused.

We gave RUEDI as well as some of the other projects that we were unable to fund at the time feedback on what we would like to see when they come back to us. I don’t know when they are going to come back to us; that is their choice.

It is very impressive, and I opened it—

I am really proud of it, yes—I would not remember all those details that you have just given us, though.

An active decision to keep the core funding has already been taken. Sir Geoffrey asked earlier about ARCHER and the curtailment of our national supercomputer. Mary Coombs is one of the facilities that will allow people to bridge between that and the new national supercomputer. Mary Coombs in itself provides more or less the same compute capacity, in petaflops, as ARCHER2; and we have a number of other distributed research infrastructures where we have already actively taken the decision to allow the community to use them in the near term to bridge to the new national supercomputer. So it is already decided that we are doing it.

Earlier I described how we have three buckets of funding. The second one is about helping to deliver on Government and societal priorities—really the industrial strategy—and the third one is helping companies to start and scale. When Hartree was set up, it was really to help companies to adopt AI and scale it into their businesses to grow their business. For sure we are going to have interventions in that space, but we have not yet fully bottomed out the portfolio that will sit within our AI programme, helping companies in that area.

I think it has huge potential.

Absolutely.

Sarah, may I add something? One of the real pluses about the bucket structure and the way UKRI and all the research councils within UKRI are working together is that I think the Hartree and Daresbury sites are ripe for additional involvement from other research councils within UKRI. Innovate, for example, is going to get much more involved in engaging with Daresbury.

Again, to bring it back to the infrastructure fund, two of the projects that we have invested in with the infrastructure fund will be based at Daresbury. They are completely different things. One is about curation of our heritage science—a project called RICHeS. The other is the national cryogenic facility, which I talked about, supporting quantum computing.

So we are putting anchors into Daresbury, because it is an important campus.

Yes, of course we think about that all the time. The most important thing is that we end up with a balanced portfolio, high value for money for the taxpayer and maximum impact. That involves us making difficult decisions. You should expect us to make difficult decisions in order to get the best impact for the public, and that is always at the heart of every decision we take.

I think we do it pretty well, actually. All the major projects that are funded across UKRI are subject to a robust approvals process. There are business cases at different levels of their lifetime. As Ian described earlier, we often seed fund a particular area to see how it is going to evolve. Then there are various gateway reviews that take place. I think we do pretty well on that front.

The Government instigated a review that Lord Willetts carried out about how you should write business cases and formulate the assessment for science specifically—for research and innovation specifically. It is very different to building a road or a rail network. David gave some really great recommendations, which are now used daily. On a daily basis, that is how we how we run our business case assessment process, and I think it has been a vast improvement.

To give some examples of that, we now ask for a 12-page business case. We were often having 400-page business cases, but they were not necessarily getting to the assurance that you need about whether this provides value for money for the taxpayer and will deliver the outcomes that we want. We have changed the delegation level for UKRI; it is £100 million now. That means that UKRI does not need to come to us to agree some of the still big but slightly smaller business cases, so you are not going through double processes. Where business cases do not work as well—David Willetts was very clear on this—is when we take too long to decide on fast-moving science and technology, and when we do not ask the right questions about what we do and do not know. We have done those two things already. We are also doing a review of the end-to-end process within DSIT. We are working closely with UKRI on how you think about—from the very early stages all the way through to monitoring and evaluation—whether it worked, whether it was a good use of funding, and what we learn from it for the future. We are doing that and we are working with UKRI on the different roles and responsibilities that we have—DSIT of course sets the overall framework and the priorities, and UKRI advises on the projects, and has the expertise and delivers them—working up through the business case process, so we both get the assurance that we need that money is being spent well and delivers the outcomes, but we do not duplicate and do not make it too slow. That is exactly what we are working on. We have made some progress. There is more to do, but we are actively working on that.

Well, better things to read because the thing is that it has to actually assure you. It has to give you the information that you need to take the right decision. That is the key. Short is good. There are always lots of annexes, but it is about having the right information to take the right decision, and to take some of the tensioning decisions we were talking about earlier.

Let me give you one example, which is at Daresbury: the cryogenic facility. That was conceived and only, I guess, six to nine months later we had approved a business case. We had taken the time to find industrial users, work out exactly what industry wanted from that platform, and gone from start to finish in six months. We can move quickly when we need to, and when the opportunity arises.

Me again, I’m afraid. Intellectually, as I said, I do not like this centrally planned approach, but let’s assume that, as you say, Sir Ian, you have now whipped it into shape and you feel you have got on top of it. Michele says you are doing it very well, so we are encouraged to hear that, but my question relates to your approach to using taxpayers’ money to take risk and develop new businesses in the cutting-edge sphere of where we are going. As Geoffrey says, I do quite a lot of EIS investment myself, so I understand that, but I also understand that what often happens with early-stage companies is that you invest quite a lot of money on several rounds, it takes a lot longer than you thought to get somewhere, and then it explodes. What tends to happen is what we saw with SoftBank buying ARM. There are actually countless examples of US companies—and others, but US companies particularly—which are more on the ball than us and basically wait for this long gestation period. They spot the growth and then they buy the business at a crucial time after a lot of the risk has been taken by our taxpayers. Given that you are doing it the way you are doing it, and you feel you are doing it well and are on top of it, if you come up with a company that is going to change things—we know that Britain is very good at innovating, and a lot of the best inventions and sports come from Britain; virtually everything good comes from Britain originally, but often we then let ourselves down by not capitalising on the actual brilliance of the idea—how can you reassure us that you are going to protect the interests of the taxpayer and let them enjoy the benefits of the risk they have taken, not give it to somebody else, which I suspect is what has happened in the past?

Perhaps I can come in on the Government aspects of this and then ask Ian to follow on. I would say that we do look very closely at the comparison with the US, and in some respects the UK is now in a much stronger position than we used to be vis-à-vis the US. When you look at venture capital as a proportion of GDP in the UK, we are now almost at US levels. That was certainly not the case 10 years ago, so that gap has closed massively. The other thing we sometimes do not notice about the US system is that it also spends a lot on public R&D. Even for a massive company such as SpaceX, much of its early technology and some of its earliest contracts were Government contracts and had Government funding. The key question is exactly as you said: how do we support those companies to go from early stage to scaling properly, and get the benefits of that in the UK? That is where we certainly have a challenge. That challenge is there in both respects that you mentioned. It is there in terms of the availability of capital: we have much less growth capital available in the UK than is in other countries. We are in a stronger position than most of Europe, but not as strong a one as the US. That is why the Government have been so focused on bringing more pension capital into that growth stage, with the British Business Bank helping to cornerstone some of that private capital, but the move there has probably been slower than we would have liked. We are starting to see some pension funds move now, but we would love to see more of that happening. They certainly have enough capital to deploy, so if that growth capital could come more if it could come from pension funds, we would be in a stronger position.

Can we avoid an intellectual debate on whether it is right for pension funds to be taking vast amounts of early-stage risk?

But they would do it on a portfolio basis.

I am more interested in capitalising on the money that you are spending.

Ian will talk a bit about some of the changes they are making in the portfolio, especially Innovate UK, to support companies on that escalator. For our part, there is a key role for Government procurement as well. Lots of the technology we are talking about has either dual uses—so in defence and security—or uses that are useful to Government for other reasons, so we should be a better customer of this technology. Through doing that, we can help these companies to grow. That is why I earlier picked out examples relating to high-performance computing and quantum computing, as those are two obvious cases where Government are a customer in the market—by no means the only customer, but a significant one. We should absolutely use our power as a customer to help UK companies to grow. We have not done that historically, and we need to do it much more.

That is a very good point, on which I think the Committee would agree with you. On Monday, we heard from Sizewell C, which is spending vast amounts of money on ageing technology coming from France rather than using modular technology developed by Rolls-Royce, so you are right that there is not a lot of joined-up thinking in Government. To your point, a lot of the things that you are investing in, some of which were mentioned by Sir Ian, rely on cheap power, as power now lies at the heart of almost every innovation. Britain has the most lunatic energy policy, which means that we have almost the most expensive energy in the world. That is why Musk is trying to put stuff nearer the sun to generate more power, more cheaply to drive all the supercomputers and development he is going to do. Do you feel that this energy policy is holding you back in what you are trying to achieve?

We are conscious that many of the companies that we want to see win have high and escalating energy costs—companies raise that with us a lot.

They are rising hourly.

That is one of the reasons we were keen to see more science and technology companies being able to benefit from the latest energy support scheme that the Department for Business launched. The scope of it allows them to benefit, and the Department for Business is currently working through that process. I think that will be helpful to many of the companies that we want to see grow. We also agreed with the Department for Energy and Net Zero, as part of the AI growth zones programme, that in parts of the country where there is renewable energy available—sometimes an excess of renewable energy is available—people who are setting up AI growth zones in those areas will see some of the benefit of that in the energy costs that they pay. We are very mindful of the energy price issue for the companies that we want to support grow. It is definitely something that we are chipping away at.

Let us not get into a philosophical debate. A lot of those renewable energy sources are heavily subsidised by the taxpayer in another way, as you know, so they pay for it in the end. The next question I had, which I think the Committee is interested in knowing, is: are you actively encouraging private sector investment alongside yours? When I have been involved in businesses, the ideas come from being involved with businesses that are innovating. You cannot sit with a clean sheet of paper. The communists found this with the USSR: you cannot sit there and centrally plan things. You need to be involved with industries that are innovating, and they will then pull you overboard. Are you actively looking—I think this is probably for you, Sir Ian—at co-investing or encouraging the Government to give businesses tax breaks? EIS is a functioning tax break, but you could give them research tax breaks or R&D tax breaks against their P&L. That arguably would then devolve rather than accrete this selection process that you are talking about. Are you looking at that?

Almost everything we do is done in consort with business. The high-level objectives that the Government set for UKRI on buckets two and three is that we will leverage three to one of private investment compared to our investment. I think there are parts of our portfolio where we will do considerably better than three to one. We also actively run schemes that are done together with companies. I will pick one: we run a scheme called Prosperity Partnerships where business has to come in alongside and match any investment that we would put into it, but they get to co-own the IP that is generated as a result of that scheme. That is done in consort. We have just run the latest scheme of that, and it was massively oversubscribed. We do that all the time.

Is that developing now? Is it a new development?

We have been running Prosperity Partnerships for many years. We are scaling them, so we are putting more and more money into that scheme and working in closer consort with business all the time.

Do you think that is a good development?

Totally. I am absolutely in favour of that. I think it is a direction of travel. I described our mission, which is to advance knowledge, improve lives and drive growth. The part on driving growth has been less overt from UKRI the past—we have been doing it, but we have been less overt about it, and I think signalling that we are a partner for business and that we want to work much more concertedly with business is a good thing for us to do.

Have you yet cracked the fusion conundrum, and are you going to deliver us this cheap energy, Sir Ian? Is that on the way?

I think these are questions for DESNZ and the UK Atomic Energy Authority rather than for me any more.

I want to follow Rupert’s questions. This follows an old prejudice of mine that we give too much basic research away to the rest of the world, who then exploit it and come back with commercial technologies that outpace our companies. I have two specific questions on that theme. First, I should declare an interest: I have a very bright IP lawyer daughter. When I go to China, they make huge play about how many IP and patent applications they are making, and I think that too many of our research institutions are not thinking enough about this. Would you agree?

Yes, I can talk about IP. The thesis of your argument is why I joined UKRI. We have this amazing remit that goes from discovery science through to starting and scaling companies. There are very few organisations in the whole world that have that remit. I think it is incumbent upon us to get better at identifying within our portfolio of investments that whole portfolio, as I described earlier, on almost all disciplines, so we are not overexposed on anything. I am talking about, across that whole portfolio, being able to say, “These are the parts of our research base that are genuinely internationally differentiated. This is the part that is world class. These are our crown jewels—and of that, this is the part that also has high growth potential.” That is such an investable proposition. To Rupert’s question, “How do you secure private sector co-investment?”, you show them that portfolio. You say, “This is the bit of our portfolio of investments”—remember that roughly half the R&D budget flows through UKRI, so we see all this. We have almost unique oversight of this whole portfolio and we can identify the really high growth potential, really high value parts of it. In concert with that, we have to have the right protections and the right commercial behaviours. I think it is worth us looking again at the IP terms and conditions that we have. We have a framework and a policy around which any investment that we make—the owner of the IP is typically not us, but the university, centre or institute in which we have invested. They own the IP, but we have a framework by which they must abide. Clearly, we indicate to them: “We expect you to exploit the IP, not just to hold it and not use it. We expect you to have the right conditions for the growth of the company and not over-stack the cap table and so on.” But we could look again at the terms and conditions that we use. UKRI’s terms and conditions are more liberal than those of other public investors—say, ARIA, the NIHR or other parts of the public sector. When they make investments in research institutes and organisations, they will have more stringent controls over IP. It is in itself perhaps a bit anomalous that different parts of the public sector have different terms. We are looking at that; we are in the process of looking to see about the right terms and conditions. We certainly do not want to be so stringent that we denude growth or inhibit company growth, but at the same time we do not want to be too liberal. There is a balance to strike. We are reviewing with the Department what the right balance is.

If you look at the Chinese universities, you see that, for virtually all their cutting-edge research, they are made to think about taking out IP and patent applications. I am not sure that the same thing is true here, and your answer indicates to me that there is more to be done.

There is more to be done.

Of course, UKRI’s terms and conditions will be really important for that, but in the current situation, where it goes to, often, the universities that get the grants, we are doing okay. In fact, UK universities generate more spinouts and patents per unit of research resource than their US counterparts—not as many, but comparatively it’s not bad. And UK IP income has grown faster than that of the US over the past decade. So there are some good things happening. However, the way universities make the most of the IP is patchy—absolutely. We have seen some real challenges where they have either kept too much or not sought to spin that out. UKRI runs a number of programmes that are about trying to encourage more entrepreneurship, more use of the research. We know from the research we have done—Irene Tracey and Andrew Williamson did a big review, looking at spinouts and commercialisation—that it is not just about the IP, although that matters. It is about, as you were saying, the desire to spin out, the thinking about how you get the support to do that, and understanding what the market is so that you actually have a market for a product rather than simply an interesting idea. There is a set of things—we had 11 recommendations from the review. We are working with UKRI to implement all of those and we are seeing some good results. We definitely have more to do, but tech transfer offices from the US and the UK are working together to agree different ways of doing tech transfer that absolutely look at the balance of taking enough IP but not too much, so that you do not put off investors, but you are getting enough of the benefit from it and seeing that come through. That guidance has gone to all the universities. More of them are using it, and we are learning from Stanford and MIT as part of that. There is definitely more to do, but there is a wider framework in which Ian’s work sits and whereby we are taking very seriously these questions. How do you do this? How do you take something from a great idea to a product for a market and do so with some of the Government contracts and with some of the support from other bodies, like the British Business Bank, and other funds, so that you can scale a company all the way through?

The second half of the question is this. If the taxpayer is putting money in—perhaps this is for UKRI and Innovate UK—why are you not thinking about taking more equity stakes? Again, it is a balance. It must not be too high; otherwise you stifle growth. But there could be an equity stake so that when the next SpaceX comes along, the taxpayer might start to recover some of the money that it has put in.

Exactly; as we think about improving the way we scale up in the UK, we are absolutely thinking about that. You can see that, with the British Business Bank, that that is exactly the approach we are looking to take as a Government. We are also looking to improve the way that UKRI and Innovate UK—and some of the amazing companies they fund—pull through, so that when a company needs greater capital than Innovate UK can or should provide, it goes through to the British Business Bank and the National Wealth Fund. So there is a way of pulling that through; we are thinking about equity stakes. It is that balance, as you say. It is about not taking so much that you put off the investors—and we know from the feedback we have had that it can do that. There are particular sectors where it is more off-putting; life sciences is one of the major challenges, for example. How do we make sure that we are benefiting from it, but not taking so much of a stake that we do not achieve the outcome, which is that we want it to scale? We have a number of institutions trying to do more on equity, and we are trying to take a more strategic approach on it as a Government; we have more to do. Then, we are trying to make sure that UKRI is part of the pipeline that thinks about it. That balance, though, is one that we are continuing to tread carefully with, so that we do not inhibit growth but actually benefit from it.

I accept that the balance is tricky, but two words that you said there worry me: “thinking about”. I think Government is very poor at “thinking about” how to get a return for the taxpayer. We had this with carbon capture and storage: the Department had not even begun to think about how it might get some money back if carbon capture and storage works. I think this is an area that needs not just thinking about, but taking proper financial advice and actually doing something about it.

Perhaps I should have said, on “thinking about”, that, actually, we are in the process of doing a massive review across Government. We are absolutely working with and taking advice from venture capitalists, from UK Government Investments, which thinks about this from a commercial perspective, and from a whole range of people. We are also bringing expertise in to talk to us, not just from the investor community but from the businesses themselves, who are telling us all the ways in which we are not supporting them to grow. Perhaps a better way to put it would have been that we are in the process of reviewing this very hard. Also, we have also taken some action already with the British Business Bank and Sovereign AI, which I think is a good example of recognising where there are problems and trying to scale things. So we are taking action and—not keeping under review; that sounds passive as well—making sure that we continue to look at what more we need to change.

Well, with this Committee as the guardian of the taxpayer’s pound, I can assure you that we will keep an eye on this, and we shall expect to see progress.

Professor Dougherty, I want to ask you about the estate. The Science and Technology Facilities Council manages a lot of the estate where research is conducted, but we know that it is currently spending only about £6 million a year managing the estate. In its Report, the NAO says that that is far below what needs to be spent. Obviously, these are very sensitive buildings with lots of quite unique facilities required; what are the consequences of not being able to keep up with what we know is now a backlog of maintenance across the estate?

It has implications on the viability of being able to run some of the great experiments that we have in the infrastructure. We are very mindful of the fact that we have ageing infrastructure. In fact, one of the changes that Ian has introduced is that there is now an estates modernisation programme. Each research council is given funding not only to upgrade the infrastructure and the estate that it has, and to make sure that we have both network and cyber-resilience, but to try to strike the right balance: “Should we be maintaining all of our infrastructure? Has some of it come close to end of life? Is it now time to look at focusing on new infrastructure?” Getting that balance right is one of the things we are working on really hard at the moment: bringing up some of the ageing infrastructure, but potentially not doing that for some of the older infrastructure that we think is getting close to end of life.

Are you actively pursuing decommissioning for some of the estate?

We are looking at costs for decommissioning at the moment, yes.

When do you expect to have finalised your costs estimate?

We have only just started work on that. I do not know how long it takes; I think three to six months is the kind of period to be able to get a good idea about that. That is really so that, going forward, we can have an understanding about what those costs will be and will really be able to plan.

Michele is referring to decommissioning a specific thing but we are constantly decommissioning buildings, largely. We have 750,000 square metres of estate, so it is a significant estate. We are constantly trying to decant people out of end-of-life buildings and facilities, many of which have all sorts of structural problems, are riddled with asbestos and so on. We are taking people out of those buildings, putting them into new buildings that are built specifically for low carbon and approaching net zero, and then decommissioning the old buildings. Basically every organisation in the world is having to do that; we are no different.

The STFC estimated in 2023 that 45% of the estate was in an unacceptable condition. Is that still the case?

Largely. We are determined to make sure that we do not have people working in unsafe buildings.

That is a good aspiration.

There are buildings that are end of life but safe. We can continue to use those with the right maintenance, but where we have HSE problems we must get people into new builds. Our priority is to identify those situations.

Is there a plan to upgrade the estate so that less than 45% is in an unacceptable condition?

Absolutely. We are going to utilise the rejuvenation and modernisation fund to do that.

Will you tell me about the plans to achieve net zero across the estate?

We need to be realistic about that across the estate because, as you mentioned in an earlier question, the STFC runs large-scale facilities on behalf of the UK. Those facilities use a lot of electricity. We use the electricity grid and that has not been decarbonised yet, so in some ways our hands are tied on that aspect. But as far as planning future infrastructure investment is concerned, net zero is central to our thinking. It is not possible to retrofit everything. We will do what we can with what we have but, going forward, we have a very clear focus on that.

Is there a plan with an end date—the point at which you expect to achieve net zero across the estate?

Yes, 2050, aligning with the Government’s targets for themselves. We are a public body; you should expect us to align to that.

Mr Mian, are you confident that UKRI is decommissioning sufficiently to enable the best value for money?

That is a good question. Decommissioning it all is controversial, but I am glad UKRI is willing to do it. I feel that it is the right choice when we are looking at buildings that are end of life and when there are opportunities to modernise. We will assess the value for money of what is happening under the performance framework that we set for UKRI: the five objectives that Ian outlined, under which we currently have 17 indicators. I do not think that we were contemplating doing something separate on decommissioning specifically.

No. The targets on net zero remain. We have quarterly meetings where we assess how UKRI is performing against the various objectives and targets that we have set. Those are opportunities to dig in deeper when we need to.

What would typically happen with a decommissioned building? Would it be repurposed for further research work or would it be a disposal?

The plots are often repurposed. The National Quantum Computing Centre was built on the plot of an old, decommissioned building, so we had to decommission the building, clean the site and then build anew on that. It is now a high-quality facility, in replacement for a low-quality one.

And it is a safe one, I hope.

I want to come back to electricity. The NAO Report mentions that one reason the University of Edinburgh was able to get the new supercomputer into operation quickly is that it had already made and been granted a grid connection. Grid connections can sometimes take many years. Do you have evidence that grid connections are a problem limiting your research and its planning?

As long as you do it with prescience and foresight, it can all be done. Your example of the national supercomputer is a great example. We have worked with our partner in Edinburgh, who have been brilliant at this, really. They have invested heavily up front in the infrastructure, the cooling water, the connections to the grid and the building, which allows us to move at pace, but you need prescience; you need to think about this in advance. The grid is constrained—I think we all know that—so we have to think about these things carefully, in good time.

Isn’t the real question, Emran, whether your upgrade of the grid is taking place quickly enough to be able to service research all around the UK? It is all very well Edinburgh having the prescience and foresight to do that, but if a new idea comes along with a particular new university and they want a large electricity supply, it is going to take them a long time to get that grid connection. Is the grid itself being upgraded quickly enough to match the pace of research that Sir Ian needs?

I think we are having to be not necessarily opportunistic but realistic about where the grid capacity is available and where universities have done the pre-work to be able to ensure that they have got a grid connection on time, and that we can use it for infrastructure. That has certainly been true on supercomputing. I think if Edinburgh had not done that pre-work, it would have been much more difficult to place that much larger system in Edinburgh. The truth of it is that it is a constraint, but it is a constraint of the same manner as some of the other constraints that we have to be mindful of when we are delivering infrastructure. Is the right space available? Is the right technical team available? Are the right partners available to make it happen? I am not aware of any issue where we have currently had to stop a piece of research infrastructure because we have not had the grid connection that we wanted.

Let me give you an example of using our assets to their fullest. This is not actually a UKRI asset, but in my previous job at the Atomic Energy Authority we used to operate JET, which is a European facility—a large facility that was very energy hungry. We had a direct connection to the grid. We ceased operations of JET in 2023, and since then the Government designated that site as the first AI growth zone, specifically because it has a connection to the grid. We are using a public asset and can build a 100 MW data centre there very comfortably. A 100 MW data centre is utterly transformative in research. You are using the national assets without having to build new infrastructure, and decommissioning a site but giving it new life at the same time. I think we just need to think strategically about the assets that we have. There are great assets in the public sector.

Thank you all very much. I think we have given you a fair examination this morning. An uncorrected version of our transcript will be available in the next few days, if there are any particular amendments that you wish to make. Thereafter, we will produce a report, as we always do, with suggestions that we hope you will all look at very carefully. Thank you again for coming today. It has been a very interesting session.