Science and Technology Committee
Oral evidence: Machinery of government changes, HC 648
Tuesday 15 September 2026
Ordered by the House of Commons to be published on 15 September 2026.
Members present: Dame Chi Onwurah (Chair); Tom Collins; Emily Darlington; Maya Ellis; George Freeman; Kit Malthouse; Freddie van Mierlo; Samantha Niblett; Martin Wrigley; Daniel Zeichner.
Questions 1-55
Witnesses
I: Dr Simon Kaye, Director of Policy and Research, Re:State; and Alex Thomas, Executive Director, Institute for Government.
II: Dr Alicia Greated, Executive Director, Campaign for Science and Engineering; and Sir Paul Nurse, President, Royal Society.
III: Dr Laura Gilbert, Senior Director of AI, Tony Blair Institute; and Dom Hallas, Executive Director, Startup Coalition.
Witnesses: Dr Simon Kaye and Alex Thomas.
Q1 Chair: Welcome to today’s meeting of the freshly renamed Science and Technology Committee for our first session of the parliamentary term. We are discussing the machinery of government changes announced before the summer recess—the abolition of DSIT—and the implications for the future of science, innovation and technology policy and decisions. We are delighted to start by welcoming two witnesses with extensive experience of previous Government reorganisations: Dr Simon Kaye, director of policy and research at Re:State, and Alex Thomas, executive director at the Institute for Government. Later this morning, we will hear from the scientific research community and the start-up and scale-up community.
I will start by asking you a very simple question: had DSIT been effective since its creation in 2023? Alex, the Institute for Government was rumoured to have been advocating for the abolition of DSIT for some time—is that true?
Alex Thomas: Thank you, Chair and Committee, for having us. That is a curveball. I was not expecting you to suggest that the Institute for Government had been advocating for the abolition of DSIT—
Chair: You can just say yes or no.
Alex Thomas: Because we haven’t. In fact, particularly given the analysis we have carried out since the abolition of the Department, we have said that it was a mistake to get rid of it. We may come on to that in a moment.
You need to make a very good case for any machinery of government change. We have done a lot of research into how disruptive and costly—not only the absolute cost, but the opportunity costs; we may come on to that—machinery of government changes are, so we generally approach any stroke of the prime ministerial pen with a machinery of government change with a bit of scepticism. When the then Prime Minister, Rishi Sunak, created DSIT, I was also personally concerned that it had the feel of a “here today, gone tomorrow” Department. Did it have a strong enough locus and sufficiently hefty policy responsibilities to establish itself in the Whitehall landscape?
I think I was wrong in that scepticism, partly because of Rishi Sunak’s personal advocacy of the Department’s agenda and partly because of the cohesive logic of bringing science, innovation and technology together. I think the Department had started to establish itself and had brought quite a lot to the Whitehall party, and on balance, we would say it was a mistake to abolish it when the new Prime Minister came into office. The fragmentation of responsibilities for science, artificial intelligence and so on across Whitehall Departments in the new set-up certainly seems to be at least potentially problematic. We may come on to that in more detail.
Q2 Chair: Thank you for clarifying the Institute for Government’s position. There has always been discussion, particularly when DSIT was created, about the trade-offs between having a central organisation that could drive something and be focused on it, but might not have the political will in each Department, and having something in many Departments that might not be able to be co-ordinated to achieve effective change. Dr Simon Kaye, what is your view?
Dr Simon Kaye: On the whole, I agree with Alex to the extent that DSIT in some ways represented the product of a lot of external engagement on the part of Whitehall, and the principal architects of the Department seemed to engage very broadly with the wider scientific and technical community while trying to arrange it. Even the sceptics who were outside the system at that point started to engage over the course of two or three years of really hard work emanating from that Department. As Alex said, an awful lot of disruption can happen. Opportunity costs are always incurred when machinery of government and structural changes occur.
Would I recreate DSIT starting from where we are now? I do not think we want to hit those opportunity costs again. More to the point, though, there is an underlying logic to the new arrangement as well. It represents the Government thinking about driving new distinctions in a really important, cross-cutting set of questions and issues for the work of Government. One is an internal versus external question: “What is the Government’s posture towards AI, technology and digital externally, versus the internal development of those things?” I think you can see a kind of departmental line being drawn between those things, rightly or wrongly.
The new structure also addresses a question of opportunity versus risk. One of the possible underlying motivations for wanting to move the question of AI development into No. 10 and closer to the Prime Minister and the Cabinet Office is that that also brings it into closer proximity to the national security apparatus. There are reasons why that expertise could be helpful across both of those things, as we have been hearing in recent days.
I think DSIT was an effective Department—an increasingly effective Department—and one that engaged effectively and broadly. But I also think we can make any system work now, and we should just try to stop moving the deckchairs around for a little while.
Q3 Daniel Zeichner: Good morning. This must have an effect on civil servants. Are people cheering, or are they thinking, “Oh, not again!”?
Alex Thomas: It is probably a bit more of the latter, partly because civil servants who have been through a few machinery of government changes know how disruptive they are. We have a bit of a theory that civil servants are almost too good at hiding the disruption of machinery of government changes from Ministers. Ministers see that the plaque on the door has changed, but they are normally not the ones grappling with either the trivial time that it takes to arrange the IT and human resources, and to harmonise all those sorts of back office-type policies, or the inevitable turf wars and sharp elbows that any kind of reorganisation brings. I think the soft disruption—or behind-the-scenes disruption—of machinery of government changes is sometimes appreciated by civil servants but perhaps not seen publicly or at the ministerial level.
There are opportunities in machinery of government changes for civil servants, so some of them will have been seeing the benefits of that. As Simon said, there is real value in elevating some of these incredibly important policy issues to the centre, and in giving them the kind of authority, direction and political focus that they deserve. Civil servants who care about that will have been enthused by it. I have to say that, on this particular machinery of government change, I have heard more concern, disappointment and scepticism than I have wild celebrations in the streets of Whitehall.
Q4 Daniel Zeichner: You mentioned some of the costs. Do we have any sense of what the actual costs are for something like this?
Alex Thomas: We did some work about seven years ago now, in 2019, on the cost of machinery of government changes. I would confess that you have to make an awful lot of assumptions about any MoG change to get to an actual number. The number that we came out with for your common or garden machinery of government change was £15 million. That was in 2019; I put that in the inflation calculator last night to see, and it would be just under £20 million now. So if you want a spurious figure, just under £20 million would be the number we would put on it.
I would approach that with particular caution for two reasons. The first is that every machinery of government change is different. This change is radically different from the merger of the Foreign Office and the Department for International Development, which would have had far, far greater costs because of the complexity of that merger.
This was some time ago, but we thought that the creation of the Department for Work and Pensions, and all the systems that needed to be addressed in that, ran into the hundreds of millions of pounds. A small moving of one function to another Department would obviously cost much less, but the numbers are likely to be higher than that kind of spurious £20 million.
The second reason is that that does not count the opportunity costs, all the distraction and the time that it takes to address the back office and human resources things, to merge grades and pay, and to deal with the new uncertainties that emerge from any set of different organisational boundaries. So it’s £20 million, if you want a figure, but I would say that it will probably end up costing more than that.
Q5 Daniel Zeichner: I have one more question for you, Alex, and then I will turn to Simon. The Government say that it will take immediate effect; that is pie in the sky, isn’t it? It will take time for people to work stuff out and work out how it is going to operate. How much time is lost as a consequence of this re-jigging of things?
Alex Thomas: As I say, there is a real opportunity cost there, and it will be different depending on the nature of different machinery of government changes and how Ministers and civil servants choose to approach it.
I do not think the Government are wrong when they say the changes will take immediate effect because, in one sense, a quick change of the brass plaque on the door does matter. Ministerial responsibilities are established either immediately or within a matter of days. Civil servants in those teams will then work to those Ministers, and the Ministers taking those decisions will be responsible for the new arrangements, so the rapid shift in responsibilities, accountabilities and decision making happens pretty much immediately.
What will not happen immediately is a change in the buildings in which those civil servants sit or the fuzzy bits around the edges, where people are arguing either that they do want responsibility for a policy or, just as often, that they do not. Working all that out takes time. The benefits of any change, and whether it brings more cohesion, as we saw with the creation of DSIT itself, accrue over time. That should be measured in years, at least, and certainly not on whether it suddenly makes things better on day one.
Q6 Daniel Zeichner: Simon, you have already hinted at some of the positives. The Government talk about rewiring the state, echoing you. What should they have done if they had wanted to do something really dramatic? Is this sufficiently dramatic to address the scale of the challenges shooting up the political agenda both here and across the world?
Dr Simon Kaye: That is a really nice question; thank you, because I think we have a bit of a reputation for being the ginger group on some of these questions. That is the contrast line for us. Everything Alex has said is absolutely right, but we would not want a Government of the day to feel paralysed in their pursuit of what they consider to be the right structure simply because of the costs we have just been hearing about.
The pursuit of the right structure should be happening. We will bemoan the fact that it was not in place earlier, rather than that we have to pay the costs now and wait the three years for the thing to be up and running. The key thing is that when we have a good structure, maybe we should try to stick with it and be smarter. I am a little bit on the fence about whether it is that kind of situation now.
What would be the big radical shift that I would want to see? If you are going to place responsibility for questions about digital government and AI at the very forefront of what you are trying to do, both in terms of transformation of the system and thinking about the way that AI is going to affect society and boost our economy, and facilitate and enable that, maybe there is a case that, at the moment, that agenda is insufficiently being held at the very centre of Government—in the Cabinet Office and No. 10—but the choice to shift aspects of it there reflects its importance.
The creation of BIST is a huge opportunity. There is the potential for a really coherent science-to-scale pipeline of sorts, where scientific insight feeds through much more quickly into the kinds of economic advantage that will make a difference to UK plc in the medium to long term. You have to do those things now to reap the benefits later.
My last thought on this is that although disruption is very important, if the personnel can be consistent over time, or at least more consistent than we have become used to in recent years, it can make a huge difference. The relationships we have been talking about do not necessarily have to die just because your email address has a different domain at the end and your address is a few doors down in the same part of Whitehall.
Q7 Samantha Niblett: The Government have said that the changes will see DSIT’s responsibilities “embedded as vital engines of growth exactly where they can have the most direct impact”. Do you think there is a risk that DSIT’s previous areas of responsibility will slip down the pecking order in their new home Departments? What should be done to prevent that from happening? It is a slightly loaded question.
I declare an interest as the chair of the APPG on diversity and inclusion in STEM and as the founder of Labour: Women in Tech. We have heard this morning that cyber-security is going to DCMS instead of the Home Office, and that the women in tech taskforce has gone to DCMS instead of Business, Innovation, Science and Trade. I will probably ask an expansive question on the back of your responses, but there is a general concern among the Committee about the deprioritisation of things that had attention because they were in DSIT and had Cabinet-level sponsorship.
Alex Thomas: The first thing to say is that the right people and the right political focus can make any lines on a Whitehall org chart work. It is important to look at these changes and assess what they mean, but it does not mean, in and of itself, that something is necessarily a higher or lower priority, although it might signal that. My concern around these changes—I hinted at this earlier—is the diffusion of responsibilities. The value that DSIT had was that it was bringing a whole set of important policy areas, decisions and trade-offs into one place. You could then have a controlling mind and, as you say, a seat around the Cabinet table that was focused particularly on those questions.
There is nothing inherently wrong, as Simon was saying, with packaging those out differently, but it means that there is a much more complicated job of work to be done across the system to bring those together and make them cohere. There are brilliant people—Patrick Vallance is an excellent chair of the AI taskforce. He has a track record of impressive work, but he has a real task in bringing together and making coherent policy on AI across Government, because it now sits in various areas. In the Government Digital Service, there was originally some concern about moving to DSIT in the Sunak move, but it had established itself and seemed to have found quite a successful home in DSIT. Moving that around again is risky. My main worry is about the diffusion of responsibilities and the grinding of gears that will be required to make that work.
A second question that is implicit in what you are asking is whether some of these really important areas will be swallowed in parts of the new Business Department, BIST. It is a big, wide-ranging Department, and it will have to balance quite a lot of different priorities. There is a real question for the Secretary of State in that Department about what they choose to prioritise, and whether some of these things get lost in it. There is sometimes a tension between catalysing innovation and making that competitive and getting out there. How will BIST reconcile those or use the size of the Department to amplify those things rather than them getting lost in the weeds?
Dr Simon Kaye: The first thing to say is that yes, I think it is important that in a big, cross-cutting Department, something like science policy is not lost down the back of the sofa. It is important to ensure that there is momentum and that the stakeholders that have been building these relationships up over time continue to be engaged properly.
If there is an aspect of all this work that I am a bit worried will become fractured by these shifting responsibilities, it is that we will have a slight reduction in the coherency of the internal picture of Government transformation. I think the responsibility for that really will start shifting now. It is not entirely clear that the Government Digital Service is now in the right place; it seems to be an entity that can do quite well wherever it is put, as you have just explained, Alex.
One of my concerns is that there is a real risk that the UK Government is now far behind the curve of the private sector—and, indeed, little bits of the public sector in this country and large parts of the public sectors in other countries—when it comes to the adoption of AI practices and innovative approaches using digital technologies. This is not a good time to take a break from pushing that agenda forward to reconfigure the way that the machine thinks about that stuff and connects it to the way it fosters external progress in those areas. That is my main question mark at this point, but I will qualify all of that by saying that it is incredibly early to be making these judgments. The dust is still settling from that shift; it could work really well.
Q8 Samantha Niblett: What you said about Patrick Vallance leading the AI taskforce is interesting. It sounds like you are saying that previously we had DSIT as a Department of its own, leading the charge and deciding what it does and does not focus on, and now the workings of DSIT have been split between different Departments, and their success will depend on a Cabinet Secretary who cares enough about that particular thing and/or an external taskforce that will basically pull the strings together. Would you say that it arguably makes the role of this particular Select Committee even more important, because what we are doing is pulling strings from across all different Departments?
Alex Thomas: It would be hard to exaggerate the importance of this Committee, and I hope that is self-evident—sorry; I should not be mischievous.
Yes, basically. Machinery of government changes about where you want the divisions to sit, or about the organisation of government, sometimes really matter, and sometimes they do not. To go back a bit, I would argue that the creation of the Department for International Trade, for good or ill, gave a Cabinet Minister a really strong incentive to sign trade deals, and that incentive might not have existed in the same way if trade had remained part of the Business Department or something else.
I go to that example because it is a really concrete one where a machinery of government change changed the incentives operating on Ministers, and it changed the political dynamic to lead to real-world changes. We might not have signed some of those trade deals if there had not been a Secretary of State relentlessly agitating to do so.
At the risk of over-extending the parallel, when there is a Secretary of State relentlessly focused on science, innovation and technology, you can say that it has created a set of incentives that are now different. It might be right or wrong, and that will depend on a Government’s priorities as to where those lines sit, but there is now more grit in the system, which goes precisely to your point. There are fewer crisply defined incentives on the leadership, both political and civil service, to draw these issues together. That is a longer way of saying, “Yes, of course. This Committee and other aspects of scrutiny are really important in drawing the threads together.”
Dr Simon Kaye: This is something I have interpreted in my analysis, as opposed to something that has come out of anyone’s mouth when talking to me, but I would say that one of the defining factors of this Government, compared with the last and the one before—I do not know whether Alex will agree—is that this Government are more suspicious of the Department as a unit of Government that can get a lot done.
They have more questions about the traditional structure of Whitehall machinery, and they are therefore going to be frequently using these taskforce-based measures to cut across departmental boundaries, or lifting parts of the Whitehall machine into other locales to create that critical distance. It is an effort to learn the lessons of things like the vaccine taskforce. One way of looking at the shift that we are describing is as an underlying suspicion or fatigue with the wider departmental machinery norms of Whitehall.
Q9 Kit Malthouse: I should declare an interest as chair of the all-party group for life sciences, and as chair of an AI start-up focused on public protection and national security.
I am amazed that you are both so relaxed in the face of what is—as far as I can see—a catastrophe for British science. What you have outlined to us is that, at a moment of extraordinary human advancement, or this elliptic curve of change, we are busy talking about the bureaucracy of departmental shifting. That is before you get on to the argument about the Treasury dividing and conquering. I think you rightly said that, now science is no one’s priority, it competes with all the other priorities that they might have.
Aside from anything else, the science community and its wider and colourful horizons now have to go looking for someone across Whitehall to engage with and talk to about what they want to do. I do not understand how this is being painted as anything other than a complete catastrophe for British science—maybe you are trying to make the best of it.
In my experience in Government, we wanted to do quite well on drugs, for example, and the drugs responsibility was across the Department of Health, the Home Office and various other bits and pieces. It was a total nightmare to get anything done, because drugs and drug deaths were not a priority in the Department of Health, so it would not focus on them at all. Do you not think the timing of this, at this moment in human development, is catastrophic? I cannot think of a better word.
Alex Thomas: Well, I have various thoughts about that. The first thing to say is that you will hear later from those who are far more expert than either of us on the science side of things.
Kit Malthouse: Isn’t it like saying, “Let’s split up the MOD while we are at war.”?
Alex Thomas: You will hear from people whose job it is to analyse and advocate for the science community. My second point is that I do think it was a mistake. I will not catastrophise it—it is your judgment about the priorities. Although I think that machinery of government changes and the organisation of Government are important—I would say that, as I work for the Institute of Government—equally, it is not a substitute for political prioritisation and drive within the system.
Were I advocating for coherence in science or AI in government above all else, I would be very concerned about this—I do not think it is the IFG’s job to take that position, because this is what it is all about—but do I think that it will be harder to reach a coherent set of decisions on science in government as a result of the abolition of DSIT? Yes I do, as I said. I think that will work against the grain of the system and I worry about the diffusion of responsibilities, particularly on AI. That is why I think the role of Patrick Vallance will be so important in driving and bringing in coherence.
We were talking about the external accountability and pressure that will operate on the system. I think that Patrick Vallance has an internal or quasi-internal responsibility to drive that through. I do have questions. The now Department for Digital, Culture, Media and Sport has been doing a real hokey-cokey over the past decade or so in its size and responsibilities. I was principal private secretary to the Cabinet Secretary when the Department gained digital responsibility. It grew that capability and itself grew. Gaining digital gave the Department a real sense of energy. If that were then removed, fine, but for parts of it to go back to what had become once again quite a small Department, and one not at the top of the Whitehall pecking order, would raise concerns, were I advocating for that.
Chair: We will come to that in more detail. Dr Simon Kaye, do you have anything to add to the specific question of whether this is a catastrophe for British science?
Dr Simon Kaye: I do not agree that it is a catastrophe. As Alex said, you will hear from people more qualified to speak on that. I come out of academia and am something of a scientist myself in a small and social way. Talking to my old colleagues and to the people I engage with in the community, they will say that it has always been a bit of a nightmare engaging with Government. I am not sure it moved that much—I think DSIT did improve over the three years of its existence—but there was an overall effort to engage. As I said, when you talk to scientists, the continuity of relationships is what makes a big difference over time to the quality of engagement with Government. I hope that that can continue under the new structure. I am a bit old-fashioned: I think that any structure can be made to work if you stick with it.
Q10 Chair: Let us come to something that has been catastrophised, in particular over the past week, which is AI and AI policy. I want us to play a little game of hunt the AI regulation, because calls for the regulation of AI have grown—this morning, the First Secretary of State will be speaking about that. We can see that AI policy, AI growth zones and AI research resource are under the Department for Business, Innovation, Science and Trade; AI strategy, the AI taskforce, the AI Security Institute and the AI Economics Institute are under the Cabinet Office; and AI adoption and preparedness is under Business, Innovation, Science and Trade. AI regulation is nowhere to be found, but regulation generally sits with the Minister for Space, Cyber and Regulatory Reform across DCMS and BIST. Do you know where AI regulation sits and where do you think it should sit?
Dr Simon Kaye: The AI Security Institute is going to be the prime driver within the machine, or it has been intended to be the prime driver for those sorts of questions for some time.
Given that we are in an unusual position of seeing a massive opportunity, where one of the principal risks associated with that opportunity is that we might let it slip through our grasp and be outcompeted by others—but at the same time, another of the massive risks is the physical or existential risks associated with widespread adoption—I agree that what we really need, as we discover the reality of this technology and its impact on society, is a highly dynamic system that can respond extremely quickly, regulate and enforce regulation dynamically and rapidly when it is called to do so, and at the same time take an overall posture of fostering and encouraging the economic prospects that might stem from this. It is a real needle to thread, an incredible balancing act and a challenge.
The dynamism, I think, is possibly one of the motivations for trying to inject so many of those aspects of AI policy back into the very centre of Government, and to create taskforce-like arrangements around them, which are considered to be capable of that higher level of dynamism. So that is my interpretation of what is going on here with that balancing act between risk and opportunity reward.
Alex Thomas: To exactly your question and the set of responsibilities, there is a huge risk of confusion and lack of coherence. This is the answer to your question on where it should sit: please let it sit somewhere and let that be coherent wherever it is. It is good that the centre of Government is taking more of an interest in this, but there is a risk to that as well. In the Cabinet Office, we have been extremely critical of the endless set of responsibilities that the Cabinet Office has accrued. I do not think it is a bad thing for the centre of Government to be driving this but not responsible for all of it. But I do think, whether it is in DCMS or anywhere else, there is a logic to DCMS, but it needs to sit clearly somewhere. I agree with all the points that Simon makes about dynamism. There is a question for Parliament as well about how we legislate on these things and how you can create a sufficiently responsive set of arrangements.
Q11 Chair: Do you think AI regulation sits in DCMS? That is where online safety and cyber-security sit. But I think you, Dr Kaye, said it should be with the AI Security Institute, which is under Minister Kanishka Narayan. So we have competing recommendations, but the key thing is that, as you said, we have not got clarity on where it is, and that is what we shall look for. Could you recommend processes to encourage or enhance departmental co-operation over AI, given the confusion? You mentioned the AI taskforce. Do you think that will address it? The regulation is not one of its responsibilities.
Alex Thomas: No, and I think you do need some kind of forcing mechanism. Joint Ministers do a very good job, although I am a little sceptical about joint Ministers. I think it is really hard for a Minister to sit between Departments, but it can work and some Ministers make it work. Whether it is the AI taskforce expanding its responsibilities or some other form of taskforce or Cabinet committee, you need a forcing mechanism. I also think that that forcing mechanism needs to come from the centre. I do not like a model where a Secretary of State marks their own homework on those sorts of things.
Dr Simon Kaye: The taskforce is going to have a really important role to play. It is right to say that the joint ministerial responsibility has to be more than simply a bridging or communicative responsibility. It has to come with the authority—I hope it does—to knit together bits of policy that need to be handled in a coherent and consistent way.
Q12 Chair: DCMS is responsible for AI adoption within Government and digital transformation, which is an example of how the Government is adopting AI. The Cabinet Office is responsible for AI adoption across the country. How can we get that better aligned or co-ordinated?
Alex Thomas: You might almost say that logically it should be the other way round, which is that the Cabinet Office, which is at the centre, should be looking internally, and DCMS should be looking out, which, again, I think is just an illustration of the policies that have fallen where they lie. It is really important that the Government thinks about this. The new Cabinet Secretary, Antonia Romeo, signalled her intent to prioritise AI adoption internally inside Government. The schemes that we have heard about over the last few years, where small pockets of the civil service have been taking on AI adoption—Redbox is one of the things they call it—have been pretty small beer.
There is a job of work to be done about AI adoption in the offices in Whitehall and their equivalents but its application to and adoption by public services also feels like an enormous agenda that needs to be gripped more tightly than it has been so far.
Chair: Another thing that has disappeared is digital ID, but we will probably come back to that.
Q13 Tom Collins: We have talked about a couple of the tensions where responsibilities are sitting across different parts of Government. We talked about AI and the growth and safety agendas. We also talked a little about how we integrate AI in the civil service and public sector. I am very interested in digital safety. Previously we had it all sitting under DSIT. Digital growth and digital safety were seen as being in tension, but at least they were in the same room. Now we have digital under DDCMS, but all the state equipment for safety is under DBIST. It should be complementary, but you have talked about how it can be seen as in tension. You have talked about taskforces doing the spanning work, but they can be scoped too narrowly. What should Government be doing to enable us to navigate this, especially given that, as Departments, DBIST and DDCMS are quite different in character and might struggle to thrash those things out together?
Dr Simon Kaye: My view—and I suspect this is the plan—is that a co-ordinating function will come out of the No. 10/Cabinet Office part of this picture that recognises that there has now been a redistribution beyond different departmental boundaries, and that will try to encourage the joining of dots. That reflects the wider pattern that we are observing of a more outward-looking and more internally convening No. 10 and Cabinet Office operation. That is presumed to be the mechanism now to make sure that those things happen, as well as the bilateral relationships between DBIST and DDCMS.
Q14 Tom Collins: That is your expectation; would it also be your recommendation?
Dr Simon Kaye: I think it can be made to work. I have not given a lot of thought to whether I would have recommended it.
Alex Thomas: Quite a lot of these organisational questions are about where you want to resolve the tension. Exactly to your point, there are circumstances where the tension is best resolved inside a Department, and there are circumstances where it is best hashed out around the metaphorical Cabinet table. What has happened here, as you say, is that where that tension is going to be addressed has moved from inside a Department to between Departments. That is not in itself inherently wrong, but the danger is that it does not get the attention that it needs and that it is actually much easier for some of those critical questions to be left unresolved because they are too difficult to resolve between Departments or between individual Ministers and their personalities. As Simon was saying, that reinforces the importance of having a forcing mechanism that means decisions in a very fast-moving area are not just left unresolved on the table.
Q15 Tom Collins: What would an effective forcing mechanism look like?
Alex Thomas: It has to come from the centre. Maybe I am loading too much on to poor Lord Vallance, but that is the beginnings of a useful mechanism. It does not have to be that; it could be a Cabinet Committee or similar that is driven from the centre—the new Office for the Prime Minister and Cabinet, I assume.
Q16 Emily Darlington: Is the Government’s focus on AI to the detriment of other areas of science?
Alex Thomas: I would be inclined to plead the fifth, or whatever the Select Committee equivalent is. I do not feel qualified to answer that. I am conscious that people appearing later this morning might be better placed to answer.
Dr Simon Kaye: I will go one step further. I have heard two things on this from people in the scientific community. The first is the ongoing concern, which we were talking about earlier, that the scientific engagement part of this picture gets crowded out by quite an understandable set of concerns, questions and attempts to grab the opportunities surrounding AI. However, there are also, very clearly, opportunities for science in the effective adoption of AI and AI-based technologies. Those two things merge with each other to some extent and might help.
Q17 Freddie van Mierlo: I want to ask about the role of DCMS. Under the previous Government, it was rumoured that there was even a proposal for it to be scrapped. It is probably fairly acceptable to say that it is one of the weaker Departments, and it is also rumoured to have a Secretary of State who is not particularly interested in it.
Emily Darlington: I don’t think that’s true.
Chair: Certainly, that is not the impression that I’ve got—
Freddie van Mierlo: It has been reported. Do you think that moving responsibility for Ofcom, digital infrastructure and other important items into that Department strengthens it, or does it risk putting those things into a silo?
Alex Thomas: I do not think DCMS has been a weaker Department in terms of the quality of its personnel or anything else—well, I will leave the political debates to you.
Freddie van Mierlo: I meant in terms of political priorities.
Alex Thomas: In terms of its status across Whitehall, the classic arguments it can have with the Treasury and so on, it is a smaller Department and has had less of a focus in that regard.
I do not think there is anything inherently wrong with putting the oversight of Ofcom into DCMS. As Simon was saying, that model can work, and there is a kind of logical coherence to it, given some of DCMS’s other responsibilities. It is a question of the focus and attention that it is given. Of all the things to worry about in the abolition of DSIT, where the oversight of Ofcom sits is lower down my priority list.
Freddie van Mierlo: Even though it now has probably the most important job of all the regulators?
Alex Thomas: Don’t misinterpret me: this is not about the importance of Ofcom as a regulator; I think it is phenomenally important. It is about the sponsorship of Ofcom shifting from DSIT to DCMS, and I am not sure that it will be done materially better or worse in DCMS than it would have been in DSIT, if I can put it that way.
Dr Simon Kaye: I have nothing to add.
Q18 Chair: I have one final question, which you may or may not be able to answer. The Government have dropped digital ID, but digital transformation plans depend on a consistent way of identifying public sector users, for example. Re:State may have done some work on this. Do you have a view as to whether we can have digital transformation without digital ID?
Dr Simon Kaye: You can certainly have digital transformation without digital ID. The advocates of digital ID point out that it would probably make a lot of things a lot easier, but all the people who were involved in the political, administrative and logistical sides of actually establishing a digital ID system in this country point out that that is not the easy bit. So it is a balance.
Chair: Great. Thank you very much for sharing your expertise with us.
Witnesses: Dr Alicia Greated and Sir Paul Nurse.
Chair: Welcome back to today’s meeting of the Science, Innovation and Technology Committee, where we are discussing the machinery of government changes announced in July, including the abolition of DSIT. I am delighted to welcome Dr Alicia Greated, the executive director of the Campaign for Science and Engineering, and Sir Paul Nurse, the president of the Royal Society. I know that George is keen to kick us off with the first question.
Q19 George Freeman: Good morning to you both. As the first Minister for Life Sciences, I used to joke that my main ambition was not to be the last, but I was actually the first of nine in about four years. As the first Minister of State at DSIT, I never envisaged that it would be abolished so quickly.
No Department is perfect, and machinery of government changes often create opportunities and challenges. What is your take on the impact of this big change? A number of us, including me as Minister, had a view that it felt very much like an AI Department, which is a huge issue, but some of the other sciences—the wet science, the harder science, the metal and the regional clusters—were a bit neglected. Is that your view, and do you see positive and negative implications?
Sir Paul Nurse: I think it was a mistake—quite a big mistake. Why do I say that? The primary driver is that science as a whole has a role throughout Government, in all Departments. It has a role not just in business, but in health, defence, transport, energy and the environment—everything. In fact, I would say that the future of our country depends on what science can put into all that we do. I was personally a big fan of DSIT, as George will probably remember, because I felt that we were on the first rung of a ladder that would recognise that and see what it could deliver.
There are ways through it, which we can maybe discuss later because it is where we are, but giving it just to Business is restricting what it can do, and that is a mistake. I would have put science and AI together, because AI is in fact science. I do not know why that is so neglected all the time. We can do better AI if it is close to science. AI is not just large language models and rushing after databases as big as you can make; it is also about promoting creativity in science and generating appropriate technologies. I would have put science and AI together to deal with that.
Practically, there are issues with this. If we talk to the Secretary of State or the Minister, I am sure that they will both say that science will be looked after properly, and I believe that they will do as much as they can. But we have to recognise that much of the activity of a Department is with the civil servants and the support. I remember BIS and BEIS, for example, which are not unrelated to this. When you talk to the civil servants, they see the big thing about their Department, which is business. Every time I had to speak to them, it was always about applications to particular business objectives, and they could not get off that. That is something for the Secretary of State and Minister to deal with, but it is a problem. It is not just the leadership, but what comes underneath. I looked it up, and the Secretary of State has nine policy areas, of which only one is to do with science. That is something we really have to deal with.
What do we do now we have it? I am sure that it is not reversible at this stage. I think that we have to identify all these problems to Government so that they are aware that these are big issues and we should do something about it. I would personally put something in the Cabinet that would be not just AI—we now have a taskforce—but science, including AI. That would be much more effective, but what would have been more effective is the Department dealing with science and AI supporting everything we do. I think that this is a mistake, unfortunately.
Q20 George Freeman: Part of the original rationale was to bring the science portfolio I had in BEIS together with the digital deep-tech AI portfolio, so it is interesting to hear your remarks. Dr Alicia, you have been a great advocate through CaSE of the broader science, research, technology, innovation and engineering.
Dr Alicia Greated: We have been, yes.
George Freeman: How do you view this change?
Dr Alicia Greated: It is a really good question—obviously, the big question. I actually went out to ask our members, knowing that I was coming here, so we have quite detailed feedback on this that we can summarise and give to you in writing. I would never disagree with Paul, because he is very important in the sector, but I will give a slightly different angle that has come from our members and from CaSE. There are different ways that you can structure any Department. In this case, there are some advantages, some opportunities and some real fears in the community. I will summarise what I think they are. I would also say, just building on what Paul said, that the noises from the new Secretary of State have been positive. We have felt very warm messages coming out from speaking to Jonathan Reynolds so far.
The opportunities are coming from the connections that are created through this new Department. The obvious one is with business. We have already heard about that alignment between basic research, coming into innovation, and the industrial strategy and economic growth, which can be positive. We also feel that there is a real opportunity in having trade in there. We have not talked much about international, but R&D associated with trade and international is a topic that I would love to cover later if we can. I know that we had comments about joint roles; on the joining up between the ministerial appointment for the Department of Health and Social Care, there are potential opportunities there. The connection that is missing, which we have not mentioned and which has come out from our members, is the connection with the Department for Education, which obviously has responsibility for universities and higher education. I do not need to tell you why that connection is extremely important.
The fears build on a lot of what Paul said, but there are two big ones for us and our members. The first one is that point about the deckchairs moving continuously. I do not need to go into that—we know how many Prime Ministers we have had in 10 years—but our members are asking for stability, whether in funding, strategy or policy. There are lots of reasons for that, but we need to be thinking more long term. We feel that a lot of the objectives of the current departmental priorities will be quite short term. Where is that long-term vision to build international partners, bring investment and gain our international standing? The second one, which builds on Paul’s point and is hopefully perceived, is that issue about the de-prioritisation of basic research. It is not just R&D; it is the skills, the infrastructure and the ideas that feed into our whole innovation pipeline. Again, I want to reiterate the support for discovery science, but also the recognition of having that long-term vision and not being short-termist.
George Freeman: I am conscious, Chair, that we have a lot of questions. I share some of the concerns, but I find myself in the strange position of actually supporting the Government’s ambition to embed science in the growth agenda, the regional devolution agenda and the clusters. We know that there is a difference between high science—innovation, research and technology—and growth around the country. I see the challenge. Chair, I know that we have a number of issues that we will unpack.
Q21 Chair: Yes. We will come back to George to help unpack them, but I want to quickly ask each of you what the most immediate challenges facing the new Secretary of State are when it comes to science. Give me a maximum of three.
Dr Alicia Greated: I would reinforce the points I have made. The first one would be that stability and long-term strategy in funding policy and the vision for R&D. The second one—I can talk about this more—is around R&D capability in the UK. I will just give the headlines now, but I am happy to go into this in more depth. We have a lot of pressures on the sector and funding changes, and our worry as CaSE is that there is no oversight of the R&D capability in the UK for our long-term objectives, not just for hitting the industrial strategy objectives. The third one is about having a really clear international strategy. We obviously support devolution, the clusters and supporting local economies, but we did not have an international strategy. Our previous international strategy was in 2019, and a lot has happened since then, so we need to be thinking internationally.
Sir Paul Nurse: I am with George. Obviously, we have to drive growth of the country; it is just that doing it this way is not ideal. What are the major challenges? First of all, the research ecosystem. I emphasise the word “ecosystem”, because it has different components. They share properties but they are different; there is not just one bucket, if you will forgive my term.
Chair: We will come back to buckets!
Sir Paul Nurse: I am sure we will. We have to have discovery research, not curiosity. Curiosity is a motive, discovery is what you are actually trying to deliver. You then translate that, which is a little more top-down, and then you apply it, which is definitely top-down. The use of the phrase top-down, and of course the opposite bottom-up, is critical here because discovery research is very important, as you rightly said, and one reason it is neglected is that it does not seem to have any objectives. That does not fit well with policymakers and maybe with Governments, but it is the way that it is done. You can read Popper, or Peter Medawar’s explanation of Popper if you want, but you do discovery research by actually wanting to understand something, with an idea. You test it, you find your idea is wrong, you change your objective, and you end up doing something else. In other words, it is a very reliable way of delivering knowledge and understanding, you just don’t know quite what the knowledge and understanding will be.
Chair: Okay, so is that your big challenge, Paul? Discovery research?
Sir Paul Nurse: That is discovery research, but as you get into translation and application, it is top-down and that is much more easily recognised. That is why we need the entire ecosystem.
Q22 Freddie van Mierlo: Sir Paul, you said the B word: bucket. I would like to go into that, if I may. What do you think these machinery of government changes leave for the bucket framework? What kind of impact do you think it will have? UKRI distributes a very large sum of money to research and with it now sitting in a different Department, what kind of impact do you think that will have on that framework?
Sir Paul Nurse: The idea of buckets is not a bad one. UKRI needed some energy, and we are seeing some energy there, so that is a good thing. What the concerns are is clarity of thought about what the buckets are and what is in them. We have work to do to more properly define what, for example, is called curiosity research—although I call it discovery research—and what is in fact a Government initiative. Is it a particular thing, like making some widget, or is it an area? We have not addressed the problems of definitions enough, if I can put it that way.
I think buckets are okay if we know what they are and what is going in them. Then there needs to be clarity of thought about how not only those things are defined, but what has happened to the budget, because we have no idea how the old budget applies to the new budget, and if you are in a change environment, that is crucial if you want to bring people with you. It is so obvious that it is something that you would be concerned about. If we are going to talk about it, I’ll talk about QR as an example—but I’ll shut up there, in case I’m saying too much.
Dr Alicia Greated: I’m wiggling my seat because I am agreeing vehemently with what Paul was saying. It reflects exactly what we have been saying as CaSE. Can I just use this to talk a little about capability, the point I raised before? With the UKRI buckets, one of our members described this as narrowing the research base. I would agree with Paul that the philosophy is focusing on outputs and value to society and the economy. We agree with that, strongly. But when you focus, you then have to narrow, and organisations—whether universities or research institutes—are then having to make choices about what they focus on.
We have got that, the buckets and choice thing, and at the same time we have got massive financial pressure on universities and research organisations, and they are making very serious choices about staffing and what departments they are going to continue to have. The combination of that, if you take it to its extreme, could mean that we lose actual disciplines in the UK. I am not saying it does mean that, but it could. We feel that no one has oversight of that—who is taking responsibility for it? If we stick to just near-term objectives—the industrial strategy, which again is important—that is fine, but as discovery research, we have to think way beyond that, as we should be thinking about 30 years down the line. We need to think about the wider context in which that sits, but I agree with all Paul’s points as well.
Q23 Freddie van Mierlo: While we have you, Dr Alicia, you will be aware that this Committee is really concerned about cuts to STFC. Is there any chance of revisiting those specific cuts to STFC with the moving of UKRI? Three of the most important UK-wide facilities—Diamond Light Source, ISIS Neutron and Muon Source, and the Central Laser Facility—are in neighbouring constituencies to mine, and all potentially face deep cuts. Is that worth revising under this new Government?
Dr Alicia Greated: I am really glad you have asked that question. I used to work in funding, and you have to make difficult decisions. We recognise that that has to happen and that we are in a constrained environment, but they need to be transparent and well-communicated—I would underline “well communicated”, because although that has improved, it has not always happened.
We focus on the physics STFC community, but the facilities that you are talking about are critical national facilities that support not only different disciplines in discovery research but small companies, big companies and innovation. As Diamond is a member of CaSE, we know that, for example, they have quite large, fixed costs, so when they have cuts, especially ongoing cuts followed by a big one, what will they cut? It will be the people.
At some point, we will get to the stage where these facilities are not operational. Our worry is that that has not been properly analysed or recognised, and we are really concerned that the facilities that you described, which are resources not just for the whole of the UK but internationally, are at risk. We are keen for this to be looked at in more depth.
Q24 Freddie van Mierlo: Do you think the Government are aware of what the impact of those closures and cuts will be?
Dr Alicia Greated: “The Government” is a big thing, and I do not know what every individual knows, but I would say that it is unlikely.
Sir Paul Nurse: I suspect they do, because it has had so much air time in the media. That is good, but also bad, because the problem is not just for STFC but across the whole board. If the money does go to STFC—I felt there should have been some solutions there—it will come from somewhere else. In that situation, you need transparency in what you are doing, so that the difficult decisions can be absorbed. It is a real difficulty if there is no transparency about how the decisions have been made.
We have to remember that the percentage of GDP—approximately 0.52%—that the UK spends on Government-supported research is well down on the OECD: last time I did the calculation, it was in the bottom half, whereas other countries like Germany spend nearly twice as much. We do not have the money in the system to support it properly—I repeat: we do not have the money there to support it properly. It is not just an STFC situation. I completely agree with everything you have said about facilities. We are overstretched.
Dr Alicia Greated: Specifically on facilities, I will not give a figure—I have a figure written down, but I am worried about giving the wrong one, so I would like to check it first—for the amount of money, but it is actually quite small when you look at Government budgets. That is something to specifically look at on facilities, but I also agree with Paul about the wider challenges.
Q25 George Freeman: To finish up on the budgets, my biggest concern is that over years of reshuffles and changes of Department, the Treasury is the big winner, because it then becomes impossible to allocate the budget. In 2021, I fought like a tiger to secure budgets, and when DSIT was created, I had to fight again. That is a real issue.
On research funding and infrastructure funding, we seem to have a problem in this country where the Treasury are normally quite quick to give Ministers capital, but not revenue. We have tended to build infrastructure, but we often do not have the revenue and expertise to support it. I inherited a couple of buildings, such as VMIC, which, without the revenue, was a big empty factory on the M4.
Is it not time we developed a better model of building specialist facilities? In life science today, you do not just need a fume hood and a laptop; you need really big infrastructure. We could do private-public funding through the catapult network and bring in industry expertise alongside academic expertise. Alicia, is the STFC crisis not an opportunity to do something more radical with the partnership with industry?
Dr Alicia Greated: I have a brief comment. It was interesting to read the responses from our members, because what came out really strongly was that the people and facilities, or bucket 4— I am not going to say what we call it—is the underpinning, forgotten bucket. We are focusing a lot on buckets 1, 2 and 3, but bucket 4, which relates to what you are talking about, is critical. That needs a lot more thought and care because it underpins everything.
Sir Paul Nurse: It is the batteries issue, isn’t it? We have a nice shiny bit of equipment and no batteries to run it. The Treasury likes that sort of spend because it is a one-off and they do not have to worry about the future, but they let us worry about the future because it is a worry for the future. Frankly, there is incompetence in that argument.
It really does matter because we invest in facilities and sometimes they are pretty good. I set up the Francis Crick Institute. It looks wonderful. Inflation ate away all the batteries and we were 20% down on our running money. I had to stop all recruitment dead for three years. That was because we did not have the batteries. We have a bright, shiny new building. We show it off to everybody who visits. It is 25% empty or we have to rent it to somebody just to make the money. This is lunacy. I repeat: it is lunacy. It is partly Government and partly private funding agencies that just do not recognise the danger of inflation in these situations.
Q26 Chair: Just to clarify, before we come to Kit, we are talking about funding, and we have talked a lot about science and hard science, like physics and so on, but science funding includes the humanities as well.
Dr Alicia Greated: One hundred per cent.
Sir Paul Nurse: And 110% from me.
Dr Alicia Greated: I cannot remember who asked it, but there was a question about whether there is too much focus on AI. I do not think this is a recent thing. The arts, humanities and social sciences are a critical part of our R&D system. We very often do not recognise that in the way that we speak and talk.
Chair: I just wanted to make sure that was recognised.
Q27 Kit Malthouse: Before I come on to my questions, is the unfolding disaster of the STFC debacle not illustrative of the problem that we are concerned about? The STFC is a subset of a subset of a subset within the Department for Business and Trade. Do we really think it is going to be a priority for the Secretary of State? It is just another one of the wasps spinning around their head. When Liz Kendall was sitting here, we were able to pin her down on it. She told the Committee there would not be any cuts at STFC, but here we are with the laser unit possibly closing. Is that not what sits at the heart of this concern?
Where is the guiding mind now on the Government’s science policy? What does the future look like, particularly given the interaction between AI, life sciences, engineering and all those other bits of science, as you said, Sir Paul? Who is making the policy now? I am confused. I do not know whether you are, too. Sir Paul, who are you going to pick the phone up to when the Crick next needs a slug of money?
Sir Paul Nurse: I do not have the answer, but we mustn’t get pessimistic. If we go to despair, we will lose it completely. Having complained—you have heard me—we have to work for a way forward and we have to work with what we have. We have a Secretary of State and Minister who are genuinely trying to do the right thing. They may be diluted and all these things—we will complain about that—but I think they genuinely want to do the right thing and we have to help them to do the right thing. Our various organisations should do that.
We also have to take the public with us. Something worth emphasising is that the UK is brilliant at science, despite the fact that it has been underfunded for years and years—all my lifetime. It is a good news story. We have amazing discoveries all the time. We do not apply them properly, and that needs to be dealt with, but let us not get complacent about how good we are. What I am saying is that we need to inspire people.
Q28 Kit Malthouse: Forgive me, but that wasn’t really my question. My question was, how is science policy going to be formulated, in your view? If you were to go in and see the Prime Minister and say, “Okay, Prime Minister, we think this is net damaging, but we are here to help you sort it out. Here is what we think you need to do to give some coherence to science policy,” what would you say?
Sir Paul Nurse: I would say this. The structure is not ideal, but if we have good people, they can make bad structures work better. We have to help them to work better in what, for me, is not a perfect structure. We need good people who will drive it, and we have to help them.
Q29 Kit Malthouse: Is it a committee, a commission or an overarching Minister? It has to be something practical—what is it?
Sir Paul Nurse: Well, Kit, it is your job in Parliament to make that decision.
Kit Malthouse: But we called you to advise us. We will be producing a report on the machinery of government with some recommendations. I am trying to elicit a recommendation from you.
Sir Paul Nurse: The recommendation is that you all take it seriously. I am saying that the structure is not right. You must take it seriously and drive it.
Dr Alicia Greated: We would be dealing with sectors. We generally deal with Jonathan Reynolds, the Minister, Chris McDonald, and the relevant civil servants. As far as we know, most of them have not changed. Those are the people we speak with directly. It is about having really strong messaging. You know how strongly I feel about public opinion. The pressure on the community has made us work much more effectively together. That is one of the benefits: when you get financial pressure, it makes people come together with a more coherent message.
Q30 Kit Malthouse: Can I make a suggestion? As you know, the Government often signal their priorities by having Cabinet Committees. Should there be a Cabinet Committee looking at science and technology?
Chair: There is.
Kit Malthouse: Should there be one that combines all these disparate bits and meets under the chairmanship of the Prime Minister?
Sir Paul Nurse: Yes, I hinted at that. We have someone dealing with AI as a Cabinet responsibility, and I said that that should be expanded to science. There are two or three positions in Cabinet that are more general, and I have in the past suggested that having some facility of that sort might give a corrective to the problems that we are discussing. That is what I meant, except that I did not have the right words, in trying to get a way forward.
Dr Alicia Greated: May I make a little comment? You started your question by saying that the STFC debacle was a reflection of what is happening more widely. I would say that we don’t really understand what actually happened in the STFC debacle. There is a question about really understanding what happened in terms of governance and decision making at that time so that we can ensure those mistakes do not happen again. That is quite a specific point. We call it a debacle, but we need to be really clear about what we are talking about.
Chair: Okay. We need to move on, Kit.
Kit Malthouse: I think George is right: some of us smell the dead hand of the Treasury in this thing. My concern is that there is no prime ministerial appreciation of the critical dependencies of things like the STFC within the science community, which need addressing. Maybe a Cabinet Committee would do that.
Q31 Martin Wrigley: I think I need to declare that I am chair of the APPG for UK innovation start-ups and SMEs. Still looking ahead, the Prime Minister says that he wants to make the UK “the world’s leading innovation nation”. How do you suggest that he goes about it?
Sir Paul Nurse: I think there is a will for that, but we have to look at several things. The first is that we have to look at non-governmental funding for innovation. Quite honestly, we are pretty pathetic. We have an invention, we apply it, we get something like Arm, it is bought out by the US, and then we lose it. Why does that happen? Why do we not invest? I am not an economist, but they tell me that the conservatism—with a small c—of investment from pension funds and so on is stopping it. That could be corrected with a small percentage of pension fund investment, because the funds are very large. Part of it is unlocking longer-term investment. That is one thing.
The second will sound strange to you. The thing about translation, which is the bit in the middle between discovery and application, is that it is mission-oriented but sort of still discovering. That needs to be controlled from the top down. You need to be ruthless with what you throw out. They drag on far too long, with time spent trying to get money into it and so on. You have to be utterly ruthless. With discovery, if you have decent people, they will discover something good. With translation, you have an objective. If you are not getting to that objective, cut it. You need efficiencies there. Those are two answers: more ruthlessness at translation and proper funding when you are applying.
Dr Alicia Greated: I agree, and I would add my previous comment about international. We are focusing a lot on our national capabilities. The UK is tiny. I have spent a lot of my life working overseas. The UK is this little country. We need to be super clear about where we have strengths and about where we want to collaborate and compete. AI is the amazing example at the moment, but there are examples right across R&D and innovation. The geopolitical context has changed so much since the last international strategy. It has been seven years. I really feel that we need to position ourselves. If we want to bring inward investment, we need to be super clear about what our offer is.
Q32 Martin Wrigley: That goes counter to the companies being bought out by the Americans. We somehow have to balance the international bit with sovereignty and sovereign interests. How do we do that?
Dr Alicia Greated: It is about us working together and understanding what we want to prioritise, which I do not think we are doing at the moment.
Chair: Indeed, the Committee’s recent report, “Science diplomacy: Sovereignty, strategy, and the global race”, emphasised exactly that point. We will publish the Government’s response soon.
Q33 Emily Darlington: As Sir Paul will remember, the Government have essentially recreated the DTI. At that time, we had David Sainsbury, a Science Minister with quite sharp elbows. He had a dedicated special adviser—namely, me—but there were other dedicated people. It was a priority for both the Prime Minister and the Chancellor. At the Treasury, there was a key unit that we worked with very closely, led by John Kingman, and we had that voice. I am going to ask you less about BIST and more about whether you think there is the same focus from the Chancellor and the Prime Minister that we had back in those Labour days, when we had the same structure.
Sir Paul Nurse: I could not agree more. We had friends in the Treasury then. You mentioned John Kingman. We have had other friends in the Treasury. I do not see them any more. That did help, particularly behind the scenes, which is really important. Somehow, we have to persuade all parts of the Government, including the Treasury, that science really matters—certainly for growth and productivity, but across everything we do. It is absolutely obvious, but it gets lost somewhere in the noise. Having friends in the Treasury, as you are implying, would really help. A Prime Minister who is committed to it—we have had Prime Ministers in the past who talked about it seriously, such as Gordon Brown—would really matter. Maybe we can persuade the present Prime Minister with one simple sentence: “If you want growth, invest in science.”
Dr Alicia Greated: I was not involved then. I am not going to say that I was too young—I was just doing a different job. The point that Paul made about public opinion and what voters care about is really important. I do not want to plug us too much, but we at CaSE have done a lot of work on public opinion, on which Florence has led. It has been really powerful for us in engaging with politicians, because ultimately, who is paying for all this and who is it for? We really need to recognise the importance of the public, the taxpayer and the voter.
Secondly, there is an opportunity around devolution, clusters and regional innovation—R&D and a lot of our anchor institutions lie at the heart of that. Again, the public are super supportive of R&D and innovation in their own local communities. That is a great opportunity for us, so we should be taking it, working together collaboratively and making the best of that opportunity.
Q34 Chair: Sir Paul, building briefly on your slogan, “If you want growth, invest in science”, we know that the investment will take some time to reap the rewards. The Prime Minister has said that he is very concerned about bread-and-butter issues such as the cost of living and the concerns of constituents on the doorstep, as it were. In a sentence for the public, not politicians, can you say why we should be spending money on science now?
Sir Paul Nurse: I would say that it is relatively small sums. When I say 0.52%, a massive increase of 20% on that would be only 0.1% of total GDP. So I would emphasise that it is not a huge sum, and it makes a difference for all the things that everybody does. They just have to look around them when they use their iPhone, listen to the radio or turn on the central heating. Everything they do is connected to science. We in the scientific community have a responsibility to identify and communicate that.
One other thing is just the beauty of it. They tell me that science always goes down well on the BBC. We really should emphasise it.
Chair: I wanted only one sentence, but I will forgive you, Sir Paul.
Dr Alicia Greated: We have a lot of data, evidence and research around this that we would be happy to share. The public are actually quite supportive of money going into R&D. The challenge is when they cannot get a GP appointment or pay for their shopping. How do you support that? The issue is that people do not know the benefits. The closer you get to someone’s individual life, the less they feel the benefits. It is not up to us to tell them, or for them to look around; it is for us to involve the public and really genuinely care.
Q35 George Freeman: To be fair to the Treasury, which is not something I am known for, they are managing a massive debt—a massive structural deficit. We are spending £500 billion a year on debt interest. To circle back to where we started, Sir Paul, I love listening to your picturesque summaries of the power of beautiful science as much as anyone, but I think we have to make the case that it is science, research, technology and innovation that are driving new industries, jobs and opportunity around the country in the clusters. I was pleased to see the Prime Minister declare an “innovation nation” as the ambition. It is a good phrase. I coined it with Boris alongside “science superpower”. But it seems to me we are trying to do it.
Emily Darlington: It was used even before that.
George Freeman: We are trying to push science into an economy that is still largely a service economy. Our public services are not very good at pulling innovation through. The City is largely controlled by overseas investors. We do not actually have many companies. We have a few, such as Rolls-Royce and others. Is there a deeper problem in our economy that we have this amazing science, but not the pillars of an innovation economy pulling it through? That is why so many companies are going overseas.
Sir Paul Nurse: I am less of an expert on innovation. I will be perfectly honest.
Chair: We will speak about that with the next panel.
Sir Paul Nurse: I would say only that we have great science, but we have to preserve it. That is definitely a risk.
Dr Alicia Greated: I just feel that what you are describing is being supported by what Innovate UK are doing, as well as our bucket structure. That is the direction we are going in. Again, I reiterate your point about all the parts of the system being critical. If one part falls over, it affects the others. Charities are included in that as well. They are all dependent on each other. We need to make sure that all those different blocks are supported and working well together.
Q36 Freddie van Mierlo: My question builds on some of what Kit said about the role of the Secretary of State and whether they will have the bandwidth to look at what the money is being spent on. When we had the previous Secretary of State for Science, Innovation and Technology before the Committee, she told us that it was not her responsibility to know what UKRI was spending its money on. They just allocate the pot for UKRI to distribute via the research councils. When you look at what some of the research councils, such as the STFC, are spending the money on, it is nationally important infrastructure—really important stuff. I mentioned Diamond, which was partly responsible for the development of the covid-19 vaccine. Is it actually the responsibility of a Secretary of State to know what that money is being spent on and ensure the continuity of nationally important infrastructure?
Sir Paul Nurse: They have to care. They can’t know every detail, but they have to care and make sure they are being advised by people who do know and do care. But it is caring that matters.
Dr Alicia Greated: I agree. They also need to care about the actual process—that it is transparent and communicated well, and that the decisions are evidence-led—but you need to be hands-off. If the Government are too close to the outputs of UKRI, that is problematic as well. It is a balance between the two.
Q37 Freddie van Mierlo: Sir Paul, just briefly, you mentioned the dangers of inflation. The previous Minister responsible for this area told the Committee that there had been no cuts to STFC. In a word, do you think there have been cuts?
Sir Paul Nurse: Yes, there have been cuts across the board because inflation has just eaten it away. I know from my experience that the MRC, which is part of UKRI, put in 1.1% inflation each year, and over four years it was 20 percentage points down. I explained at the time that it would happen.
Dr Alicia Greated: CaSE has challenged that terminology. It is interesting. We would probably use the word “cuts”, but what is the difference, really? Ultimately, it is what people feel on the ground if jobs are being lost.
Freddie van Mierlo: The difference is political cover.
Dr Alicia Greated: Yes, that is true.
Chair: I think Freddie is right, in that the difference is accountability and being able to set out what has happened to the science community, the general public and Parliament. Thank you very much for joining us.
When the Secretary of State was here last time, we agreed to move away from the “buckets” language. We seem to have gone back there, for want of alternatives. But whether it is “buckets”, “funding streams” or whatever, it is clear that this Committee needs to be able to scrutinise it effectively and that our Government need to be accountable for how they are delivering on the science agenda. Thank you very much for sharing your expertise.
Sir Paul Nurse: Thank you for listening to us.
Dr Alicia Greated: Thank you.
Witnesses: Dr Laura Gilbert and Dom Hallas.
Chair: Welcome back to the final part of our session, in which we are discussing the machinery of government changes announced in July, particularly the abolition of DSIT. I am delighted to welcome Dr Laura Gilbert, the senior director of AI at the Tony Blair Institute, and Dom Hallas, executive director of the Startup Coalition.
Q38 Maya Ellis: In your own way, you have both been very critical of the closure of DSIT. As an open question, what are your main concerns?
Dr Laura Gilbert: It is not particularly the closure of DSIT that I am critical of; it is the mechanism by which those operations take place. We have seen, over and again, new Governments coming in and very quickly announcing structural changes. I am sure you are all familiar with Chesterton’s fence—the idea that you should not break something without knowing why it is there.
If you look at professionals in the space of reorganising businesses, they never come in on day one and just restructure everything, because you are breaking things without really understanding what the knock-on effects will be—particularly when you are doing that to solve a problem you perceive as having been caused by the previous Government making a structural change. It felt quite on the nose, really.
There are a number of ways that you could structure digital and tech in Government, and I have views on how that is best done. However, the process by which these changes take place is actively quite harmful, and that is not particular to these particular machinery of government changes but about how the process works. The communication is often confusing and the people impacted have no warning to prepare for it—that is true inside Government, and also in the public, commercial and business spaces. The way humans are treated when a machinery of government change happens is very damaging, and there are quite predictable knock-on effects.
Having been in the system for the last one, I know that the message received is that the powers that be—in the civil service and the political side of Government—do not really care about you as people. They have not given you any thought. They think you are fungible: it does not matter where you are, what you are doing or what your skills are—you are just a piece to be moved. In this particular instance, a couple of years after the last one, the team I was running during the last MOG still had not got the new laptops for the new Department they had been moved into by the time they were moved into another one.
Realistically, what happens is you take a bunch of teams doing something—sometimes really well, and sometimes not so well—and you say to them, “Right, we are going to put you in a different building; we haven’t decided which one. You are going to have a new email address; we haven’t set up the servers yet. There is going to be a new office somewhere, and it is definitely going to be a bun fight over who can get themselves into the best office. We’re not totally sure what your remit is; there is probably a hiring freeze, so you’re not going to be able to backfill roles, and you do not know who your political sponsor is any more.” That takes months, up to two years, before it washes through the system.
The effects are very predictable. What happens? People become dissatisfied and start looking for jobs in other Departments or outside. Of course, the best people find those jobs very quickly, so you are preferentially moving people out of the system who you would probably like to keep. It is frustrating to see that happen time and again.
That is my criticism. It is not this particular change to the way Government operate but the process itself, and there are ways to do that well.
Q39 Maya Ellis: Thank you for sharing that. I appreciate there is that broader picture. Can you elaborate a little on that? Looking at DSIT itself, do you have specific concerns about how it will change and the particular impacts it might have, given everything you have just said?
Dr Laura Gilbert: At the moment, we separate policy and delivery, which is always my main concern. When you take the people who know how to do the thing away from the people deciding how the thing ought to operate, that is a meaningful barrier to good outcomes. It is not that you could not put machinery in place that reconnects those two things, but there is a recurring organisational cost that comes with separating two things and not having them together.
In my personal opinion—I think I said this last time I was here—a lot of what could be improved about Government would come from having technical people who understand how to build things sitting right next to those working on policy and procurement, and advising them directly, “How much would it cost us to do this ourselves? How long would it take to build? Therefore, if we are procuring, what is a reasonable cost?”
The same applies in the policy space: “What would be the real knock-on effect on engineers and businesses if we set the regulations this way?” When you separate those two groups of people into two teams with no line management in common—except for one excellent AI Minister, in many cases—that is a real problem, and it is a cost that does not go away.
I have real issues with that. You have also moved the part that interfaces with industry broadly into DCMS, which does not have the same historical links with that kind of industry. You are looking to that Department to reframe dramatically if you want it to succeed and asking industry to work out how to connect with a completely new organisation, without, thus far, any guidance on how to do that.
There are a number of other reasons why I do not think that science, tech, AI delivery, etc, should be separated, but there is a real point about looking to the future as well. You have taken the frontier technology away from the people who have the best experience of going through frontier technology transformation recently. We have a new AI taskforce in the Cabinet Office, which Patrick Vallance will chair. He is excellent, but as I read it, they do not have any analytical capability. I do not believe in taskforces that do not have analytical capability; they then have to commission other people to do it. Again, you have another structural impediment that will slow you down repeatedly from good delivery. I could go on.
Q40 Maya Ellis: That is really helpful, thank you. Dom, I have a similar question for you. We had some lovely comments earlier from Dr Paul Nurse about looking to the future. Could you summarise your concerns, but also tell us whether in these first few weeks you have seen some shoots? Are you optimistic about how it could work?
Dom Hallas: First of all, other than that, Laura thinks it is great. [Laughter.] At the time, we, along with our compatriots or frenemies—however you want to describe it—at Tech UK, came out and said we thought it was the wrong change at the wrong time.
Laura has very effectively covered off the machinery of government stuff. I can talk a little more about the policy and how our community sees it. There are three areas where we see the policy impact most clearly. One is the risk of crowding out big questions in a much bigger BIST Department. The truth is that the Business Secretary covers an awful lot of ground and always has done. Adding in even more ground means that it is even harder to cover.
I was a civil servant, and I know just how much that Department ends up covering—steel bail-outs and everything across a whole spectrum of issues. Understandably, a lot of those things are very political salient and high priority, and in those moments, they need an awful lot of political cover. You end up boxing and coxing, and moving around to try to cover as much as you can, and putting out fires, fundamentally. The fear in my mind is that amid the account management job that the Business Secretary has to do with really important business stakeholders across the spectrum, the science and tech community, especially the start-ups and scale-ups that we represent, might get squeezed out a little bit.
On the green shoots, to be fair to Jonny, the wider team and the wider governmental structure, they have made an effort in this first period to address that and make sure it is not the case. I have been relatively encouraged by that, but it can quite clearly be a changing picture depending on exactly what fire suddenly starts and has to be put out.
The second structural challenge from a policy perspective is on regulation. We now have a situation where it is potentially split across multiple Departments. If you take AI, you could have a situation where the harms or online safety aspects sit at DCMS, and you have a taskforce that is ostensibly responsible for a lot of those conversations. If you look at the discussions today, which I am sure everyone in this room is tracking, around what is happening at the frontier and potential regulation, people are looking to the AI Minister, who is directly involved in the taskforce, to see what he is going to say. At the same time, you have a BIST Department that is going to think about the regulation of that in an industrial setting, or the wider regulatory innovation office, or how it impacts other sectoral regulators. You have this split function.
Frankly, we have interacted with all those Departments over different periods, and they have different views because of the stakeholders they serve. There is a genuine tension—or at least pros and cons—of whether you want to resolve that within one Department or whether you ultimately want to have people fight it out in Whitehall. I do not think there is necessarily a right answer to that, but from our community’s perspective, you can end up with a much more chaotic landscape.
The third thing—Laura touched on this, but I will build on it—is about the use of technology in Government. Again, you have this split between a lot of the GDS-type functions that now sit at DCMS and the AI taskforce that sits at the Cabinet Office and whose job it is to spread AI across the Government. Paradoxically, I hope they will collaborate really closely; that is ideally what you want, but there is a risk that you will have the sexy end of the technology and AI at the Cabinet Office with all the cool, sexy people, and the bread-and-butter technology being delivered at DCMS. Frankly, if I am an official leading some of the work at GDS, I don’t know what that tells me about my work, on a very base level. As Laura rightly said, there are some very talented people who oftentimes could get paid a lot more in the private sector who are looking at themselves now and thinking, “Hang on—I’ve been moved from pillar to post.”
At a base level, though, we are really keen to get this right. At the end of the day, there is no point in crying over spilt milk. Tech founders and the community we represent just get up and get on with it. I do not want to dwell too much on the pure machinery of government, but from our community’s perspective, it is worth reflecting on what this means and looks like to outsiders.
Everyone in this room is very familiar with the idea that Government Departments move around and someone is going to sit over there now, or there are complete changes, but when our founders heard about this, their question was, “Is this work going to exist any more? Do these people get sacked? What happens?” They do not understand. It is not their job to follow politics every day. That is before the answers to who exactly is sitting where eventually came through, maybe after everyone got their email addresses, as Laura said. I have also been through that process—it wasn’t fun.
We have a version of this conversation that we need to have about how you think about Government functioning, and then you have a version that other people, including our company, perceive from the outside, which is even more challenging from their perspective, and even more opaque.
Every structure has pros and cons. There are huge opportunities with the structure that has now been created, when you look at the BIST landscape and that journey from R&D funding to business finance to growth. There are clear opportunities in the structure as well—there is no right answer to this question—but it is important now that we understand what the opportunities of the structure as it is today are, and how we make sure that we therefore take those opportunities.
Q41 Maya Ellis: Picking up on the second point you made about the different views within Whitehall and who will arbitrate on that, linking it to the people you represent, who do you think should arbitrate and, realistically, who will end up arbitrating?
Dom Hallas: You are getting into who makes the decisions in Whitehall—if I knew that, I would be much better at my job. In reality, every policy discussion within the Government ends up being about winning coalitions of people in Government. I see our work at the start of coalitions split into the work we do on, broadly, economic policy, which is the tax and business landscape, and then the technology landscape and wider regulation.
On the economic policy landscape, it tends to be that it does not matter how many ideas outside Departments have; the Treasury ends up not ruling the roost but certainly having a pretty strong say. Oftentimes, when you think about the potential impact in some of these areas, it is fair to say that on the economic policy side, BIST is going to be really important but you are clearly going to have really strong input from Treasury and elsewhere.
The challenge on the technology policy side—this is an important question for capacity, not just outside Parliament but at the centre as well—is who is going to make those decisions about some of the big questions. We are here on a day when, clearly, we are having deep conversations about AI regulation. The God’s honest truth is that I am not 100% sure who is driving that conversation from the centre at the moment. There are definitely people working on it—no doubt about it—but that decision has to be taken politically, at a fundamental level. People voted for someone to take that decision. The honest answer is that it is an open question and it will change issue to issue, but I am sure that the answer will be very important when hopefully both the people in this room and, frankly, I discover that.
Q42 Chair: I am going to ask each of you a quick question, which I think might help the rest of our discussions. Where do you think the real deep understanding of AI is in Government now? In which sub-departmental group or Department?
Dr Laura Gilbert: I think it is fragmented at this point. I come from a very biased position, because of course I am an engineer and I built the engineering team.
Chair: Me too; I like those biases.
Dr Laura Gilbert: Quite. I think it sits with the engineering capability in ACE, in i.AI, the incubator, and in GDS—those sorts of engineering teams. I think they have the deepest understanding of the technology, because they actually build it. There is huge expertise in the policy teams. One of the things about DSIT was that that was knitting together over time—they really were starting to learn from each other. There is possibly an advantage in spreading that out so that the learning propagates, but I would look to the technology teams first and foremost.
Q43 Chair: Dom, in your engagement, where have you found the best understanding of AI?
Dom Hallas: I agree with Laura to some extent. When you look at the technical teams, there is certainly really important capability there. When it comes to the policy perspective, the honest answer is that it is spread in disparate ways, and you find individuals who are really talented and have a deep understanding. Crucially, that is not, “They are super senior and they have a good understanding”—it is quite the opposite; you have a real mix of talent. In some ways the strength of DSIT as a concept was precisely that you were trying to draw as many of those people together and therefore have a cluster of talent that would then uplift, and the rising tide would lift the wider boat.
Clearly, if you look at the taskforce and the people who are being drawn together—we hosted dinner yesterday with Ollie, who is running the taskforce—it is obviously an extremely talented set of civil servants who are working on these questions. They have thought about it deeply and are at the top of the tree. Frankly, if you are running, like I said, a sexy Government taskforce out of the Cabinet Office, you are going to attract incredibly good talent.
The question I have, and the thing that matters the most from your question at some level, is: will some of the more disparate questions when it comes to AI and the impact on both the economic landscape—our companies—and on Government services be able to attract the same level of talent as some of the core functions, or will it just be split out, as it is at the moment, and we will therefore have a lower quality of work overall?
Chair: It is about the water cooler conversation impact not happening any more in DSIT, together with attracting the best.
Q44 Samantha Niblett: It was really interesting listening to what you were saying about the impact on morale and dispersing the expertise. I know that there was a whole community that sighed sadly when DSIT was disbanded because of the potential impact on the women in tech taskforce. My understanding is that that has shifted to DCMS, which seems a bit strange, for various reasons—Emily and I briefly discussed that yesterday. How much of an impact can the external organisations have—particularly within your network and the Startup Coalition, Dom—on insisting that the Government make sure that they continue to prioritise it?
As the founder of Labour: Women in Tech and the chair of the diversity in STEM APPG, I have written to the Prime Minister saying, “Can we please make sure that this particular taskforce has a Cabinet-level Minister, as it did in Liz Kendall?” Obviously, it was her creation—her baby. How can external organisations keep the pressure up so that this does not lose its impact? I know that there were civil servants who were working on this project, and I wonder what is happening to them.
Dom Hallas: We have had direct conversations about that with some folks in Government. My former colleague Vinous—she left on Friday—sat on the taskforce on our behalf, and I know that in the immediate post-restructure conversations it was a case of exactly as you say: “To what extent do we want it to sit in DCMS or would it be more effective spread across Departments in the wider way in which the technology community is?” My understanding is that, for various reasons in Government that we do not quite appreciate, people want it where it is.
If you step back and look at the way in which broadly the machinery of government structure has segmented, that means that it is within a digital inclusion brief. Frankly, from a personal perspective, I think that is completely irrational. We have to be clear that the role of women in technology is not an inclusion question alone. The fact that women cannot raise venture capital really matters, and the teams that are working on it, as good as they are in the inclusion brief within DCMS, are not responsible for those things.
We have had positive conversations with Ministers about this, to be fair. My hope is that there is a deep understanding elsewhere in Government, at BIST more generally and other places, that will take on the constituent work that the taskforce would have been doing, and obviously there is natural cross-collaboration in Government. That is a really good example of something that has been split because the Department has been split, and I am not sure it will have the same impact in its more fragmented form as it perhaps could have done.
Dr Laura Gilbert: I completely agree with everything that Dom has said. One of the things that we sometimes forget about when we are discussing these kinds of situations is that this is a very human issue, and it has been driven by individuals and the incentivisation around them, and we need to be a bit realistic about what that looks like. When you are performing a change of this magnitude, the people responsible for it have an awful lot to deal with. There are things at the bottom of the pile that do not get handled, and it took a while to get them in in the first place. It is not that the Government are broadly, as a unit, dropping the ball on something; it is more that individuals who are tasked with this have a set of priorities—they are probably very busy and pretty tired, and are under a lot of scrutiny—and it just does not make it on to the priority list.
It is about a couple of things. One of them is public pressure—drawing light on it and getting enough people to weigh in to make a difference. The other thing is thinking very carefully about the timing and when the right time to draw people’s attention is. Somewhere between the immediate panic and before everything is settled is the right time to do that. It is about bringing attention to the things that really matter. This is certainly not the only thing that has been dropped; there are a bunch of advisory boards that do not know where they sit or whether they have impact any more. There are lots of people in the civil service—not just with this MoG; I have seen it with previous MoG changes—who hold something with great passion and have worked very hard to get it to the place that it is and all of a sudden, it is gone.
Dom Hallas: One thing that is really important is the distinction between things that are machinery of government and restructure questions and those that are ministerial change questions. I think that the split part is a good example of a machinery of government question, whereas a ministerial reprioritisation thing like the stacking the pile that was mentioned can happen as well. Understanding the distinction matters too.
Q45 George Freeman: You have both been very eloquent on the scale of the challenge. You are veterans of Whitehall and policymaking. As the dust settles, what do you think is the genuine shift in policy and priority where you would be saying to international investors and start-ups here, that for all the noise and fury, what has happened is—it looks to me—a big ambition on AI that is fragmented, and science backing business. You are the experts, so what do you think is the big takeaway of where the Government are really prioritising?
Dr Laura Gilbert: So, you’re asking for the sales pitch.
George Freeman: No, not at all. I am asking for your advice based on seeing a few of these and what you think is likely to change—for the better or worse; there might be a bit of both—through all this.
Dr Laura Gilbert: There’s “What will change?” and “What do we want to change?” and the point is to push to what we want to change. A good outcome would be revamping procurement that particularly supports small businesses to succeed, proliferate and grow quickly with the right kind of backing. It would look like much more effective procurement so that we spend the public’s money much better and on the right things, so we do not have to abandon them after four years. It would look like an international signal that the UK intends to lead, particularly in the adoption of technology. There is a huge conversation in the policy space—I am quite AI-focused—at the moment about safety and risk, and whether there is a space for the United Kingdom to be at the front of that.
On the investment in science and technology, we have a very strong university sector and a pipeline into business that stays rooted in the UK is something that we could really work to expand. Attracting engineering talent in frontier technology from other countries, and making this a good place to grow a business, and to live and work, is what I would want to see out of it.
Dom Hallas: If I were to steelman the case for the restructure, I think there is a clear throughline at BIST in the way that you have the research funding to the public financing landscape and the scale that, in theory, if it works well, can work really well. If you look at the AI policy structure, there is an argument—we saw this in the past two years—that there are tensions and trade-offs in what the Government might need to do on AI. Energy is a good example: over the past two years, the Energy Department wheeled in and said, “Well, no, that’s not going to happen.” Having a taskforce at the centre that has prime ministerial oversight and drive could maybe smash through some of those barriers.
If you talk to some people in the system, that is in theory the way in which that could operate. Personally, I am more sceptical about whether that would be the case, just because I think that some of those trade-offs are inherent. We have to acknowledge that some of that is about political will and genuinely political choices, whether or not they are the right or wrong one—obviously, I have a view in my job.
So much of what my community and international investors, certainly VCs, care about is what I loosely characterise as vibes. Unfortunately, getting rid of a Government Department that has science and technology on the tin gives bad vibes, but what the Government can do and have been trying to do—we have been encouraged by this—is elevate people that the community I represent recognise as talented, effective and understanding of their work. Jonny has done a really good job of that and going and doing the outreach. Kanishka, with his understanding of AI more generally, and certainly in the community, has a lot of respect.
I think that as much as could be done within the framework of a restructure is being done. The point—to go back to the point that probably Laura and I started with—is that I am not sure it was necessary to begin with.
Chair: Thank you—very good question. I guess a vibe-coding Government need to make better progress.
Q46 Emily Darlington: I am going to read this, because I get really confused. The Cabinet Office now has responsibility for public sector AI adoption. DCMS has been tasked with driving “the modernisation of government and public services via the GDS”—the Government Digital Service—and the Office for the Prime Minister and Cabinet will pursue opportunities for “public sector transformation”. Where does this leave the public sector transformation agenda?
Dom Hallas: Laura?
Dr Laura Gilbert: No! [Laughter.] I do not feel it is entirely clear; I am not certain that it is completely settled. I do not feel clear about who is actually going to be driving this in real terms. Do you, Dom?
Dom Hallas: I think that it is worth stepping back. When you think about the history of machinery of government change, when it comes to science and technology policy broadly, you had the creation of DSIT as a Department under the last Government, pre-election, and separately from that, when Starmer won the election, you had the movement of GDS into it as part of the restructure. There are certainly civil servants within the system who want to take very strong ownership of this agenda. They think it is really important. You can see that in part of the AI taskforce work. It is quite clear; the Cabinet Secretary has been very clear that she sees this as really important, and that is why she wants to have more of a grip of it at the centre. That is really encouraging. It is really important that we do that effectively.
I think the question is how. When I worked in Whitehall, I worked at the Department for Exiting the European Union, which was set up intentionally to serve a task, but what happened was that in fact it was deeply underpowered. It was, frankly, outmuscled by larger Departments. It was a bloody nightmare, to be honest. Quite candidly, the challenge is that if you do not have Departments that are capable of arguing and driving some of that change, you have a bit more chaos. That chaos still exists for the time being; that uncertainty still exists. I am sure that this will bed in, and I think that there will be an answer to the question that you have asked in the coming months, partly just out of the rules of the road being set by default as people are working and operating within the system. But that is probably the area where, to my mind, it is least logically structured as currently constituted.
Q47 Emily Darlington: I read those reports about Antonia Romeo wanting to seize it. My question is: do you have any insight into why GDS is at DCMS and not in the Cabinet Office?
Dr Laura Gilbert: I was quite surprised by that one. There is a bit of a fray when these things happen. Again, it is the human thing of people with competing interests making arguments for what they want, both on the political side and on the senior civil service side. It is difficult to know whether somebody won that fight and therefore won GDS or whether it was more incidental and it was played off against something else. I was genuinely quite surprised by it. You can see an overarching argument for putting delivery in one place, but I think GDS was significantly disempowered initially when it was moved into a new Department from right at the centre of Government, and I was surprised to see it being moved further away, given how important digital delivery is to the Government’s agenda.
Q48 Emily Darlington: I am equally surprised, because those detailed conversations tend to happen from perm sec to perm sec with the Cabinet Secretary. She would have been involved in those discussions, so I am a bit confused. That tends not to be a political decision; it tends to be done at that level of Government. I do not understand why that happened.
Dr Laura Gilbert: It certainly can be political. Obviously, the GDS move happened at a political transition. I would like to understand the access talks more. That is where these decisions take place. In my experience, they are very narrow and do not tend to take into account expert leadership. Certainly, with the original transition of GDS into DSIT, as far as I am aware, there was not a conversation at the level of the access talks with the senior leadership of those teams that were actually moved. I would like to see that happen, because then they could say, “Well, here are the pros and cons.” This is the point about divorcing the policy or the strategy from delivery and the people doing the delivery. It happens in so many different ways in Government when policy is built generally: the people who actually have to make it happen do not get consulted and do not get to say, “This will be the knock-on effect.”
Q49 Emily Darlington: You said that alongside the women in tech taskforce, a lot of advisory committees were up in the air. Which ones do you think are priorities that we should make sure land somewhere?
Dr Laura Gilbert: I am coming from bias again, because I am on the GDS responsible AI panel. The question that came up to us immediately after the whole thing was announced was, “Do we go with GDS, or is it more of a policy advisory function?” I do not have a list of all those advisory committees. I could certainly go and find them.
Dom Hallas: The advisory committee that I was on got abolished. I was chairing the expert group on the National Data Library. From an advisory perspective, it might be interesting and helpful for the Government to consider bringing back what used to be called the Digital Economy Council to help to paper over the cracks in governance and the structure of technology. To be totally honest, I think the framing may be a little old-fashioned. We have a digital economy now, frankly, but I certainly think it could be important within the Whitehall machine to bring together in one room the industry leadership, the wider civil society perspective on technology and the relevant senior politicians responsible across the different Departments. That could be a good opportunity to iron out some of the tensions we talked about within a structure.
Emily Darlington: That is quite interesting, and that is in line with what Sir Paul Nurse said.
Q50 Maya Ellis: I have a quick question about where we are on regional innovation and devolution. In the bigger picture, given all these changes, we know that local growth is in MHCLG now, so where do you feel those issues are? I am particularly interested in non-mayoral areas like Lancashire, where I am. We have a top 10 research university, Lancaster University, where we have now lost about 200 jobs in the innovation roll-out, because all the ERDF and subsequently SPF has gone. Where should the Government be heading now to ensure that we do not lose those innovation ecosystems within the whole devolution agenda?
Dom Hallas: The tying up of the business financing landscape that sits at BIST with the science funding and R&D landscape is a good opportunity. We speak to businesses all over the country every day, and there has clearly been a tension: they might get grant funding or loan funding support from, say, Innovate UK, but that does not necessarily translate to private financing once they have had the opportunity to build something and they have a functional product, and what does that mean? Quite clearly, that tension exists more deeply outside London and the golden triangle, where private venture capital is less abundant. I think having it all sitting at BIST is a good opportunity to understand that landscape and context more effectively, working with MHCLG and other Departments, so there is definitely more that can be done there.
The other area on the devolution side where there is potentially an interesting opportunity—this ties in with more things sitting in the Business Department—is in understanding some of the questions about the taskforce and thinking about AI infrastructure development. How do we have a more candid, devolved conversation about those things that acknowledges that communities have to benefit from stuff, frankly? When you have more of those discussions in one place, and when you are able to co-ordinate more effectively in theory, that should work more effectively. I think there are opportunities there, and I think the structure could work better, but we will have to wait a little to see how it plays out.
Dr Laura Gilbert: There is something about shared learning. We could really power up the idea of sharing learnings and sharing things to do with procurement processes. I will not go on about it, but there is space for a lot more open-source delivery.
Chair: Ah!
Dr Laura Gilbert: Yes, I won’t get into it too much, but it is my big thing. There could be a lot more open-source solutions requiring a lot more open source. That way, you build it in one place, and when you translate it to a different area, a local business there can pick it up, tweak it so that it meets local needs and then quickly deploy it. That saves money for local authorities and drives local businesses.
Chair: Thank you very much. I will bring in Tom now, but I like the reference to open source; you can never talk too much about that.
Q51 Tom Collins: You have mentioned a few times that we have the new AI taskforce, which will be chaired by Lord Vallance and headed by Minister Narayan, who will also attend Cabinet. You have talked a little bit already about whether that will be internally focused or externally focused. Previously, Sir Paul Nurse talked about possibly expanding that scope to bring in science as well, so that science and AI are complementary and paired. What would you like to see that taskforce focusing on?
Dr Laura Gilbert: I think we are missing something more broadly at the moment, which is a delivery board for science, technology, AI and so on. That could be the basis of it. What we have done most recently is really move various parts of science and tech into pre-existing silos. It is very difficult to get out of silos, and that is for reasons of culture—or the water cooler. Also, each Department is on a different technology system, and they are not interoperable.
My inclination would be to power that up pretty significantly to give it a strong remit, with outcome objectives that it is strongly tasked to meet. I would definitely put that analytical and tech capability into it—that is, people who really know how to build the thing, so that it has really good advice. I would try to turn that into a delivery and oversight body over the whole piece to draw those threads back together again. I would give them an internal tech budget so that everyone can be on the same Slack channels and actually collaborate properly.
Tom Collins: Brilliant; I really like that. Thank you.
Dom Hallas: I am a little bit squeamish about the idea of almost looking at the structure that has now been built and saying, “Okay, how can we make it more like the last one?” I think there is a risk, quite candidly, that you will lose the benefits of the existing structure while not necessarily gaining the benefits of the previous one.
Far be it for me to disagree with Paul, as eminent as he is, but I think that in reality, trying to pile more stuff into a taskforce that is intentionally super narrow—I think it will have 50 staff or something—is a recipe for a-bit-of-everything-ism. In reality, I think it is much more about asking, “What are the core questions in AI when it comes to the state, our overall domestic capacity, our sovereign capacity or however you want to frame it? How do we smash through as many of those barriers as possible?”
To Laura’s point, the question of what a team or the taskforce will then require to do that work is really interesting. Maybe it does require an analytical capacity that it does not currently have, or maybe it does require more technical capacity. I think that it is about really clearly focusing on those core tasks, as opposed to trying to rebuild a science and technology Ministry but just calling it a taskforce, which would be a bit of a hiding to nothing.
Tom Collins: So we should lean into the agility of that taskforce, rather than trying to expand it.
Dom Hallas: Yes. This goes to Emily’s previous question about the rationale of where GDS sits today in DCMS and elsewhere. I think that we have to be clear that the very intention of setting up the taskforce in the way that it has been done is to give it agility, and you can look at the GDS decision in that context. GDS is a big delivery function, fundamentally, and that is not necessarily what has been moved into the Cabinet Office.
Again, you can have a much wider argument about whether you could have had a taskforce and then still have had GDS in the Cabinet Office, but separately. However, where would it have sat ministerially? Would it have sat with the Minister you have just put in the Cabinet Office? There is all that sort of stuff’; we can all play “Football Manager” when it comes to the structure.
In reality, from my perspective at least, it is about how we lean into the agility, understand that this is a thing where pace really matters in a world where pace really matters, and then leverage that in the most effective way we can.
Dr Laura Gilbert: It is a really good point. In Government, we have a few very good taskforces driving meaningful change, like the covid taskforce on vaccine delivery; we have also seen a lot of taskforces whose names we cannot remember, but which were definitely there for quite long periods of time. I think the difference between those two groups is the outcomes focus.
You have mentioned the need for urgency. They will need to be very clear about the actual thing that you are delivering and how you will know when it is done. That is quite difficult with something as generic as AI.
Q52 Tom Collins: You have talked about keeping the scope sharp and about being really clear on outcomes and purpose. Are there any other important emerging technologies or areas—you have mentioned silos already—that you think are very important but that might miss out as a result of this area of focus and the priority it has been given?
Dr Laura Gilbert: It is a difficult one. There are lots of different areas of technology. Quantum is obviously coming up; there is an awful lot in the cyber space. There will be new technologies coming that we do not know about, and they will be coming fast. I worry a little when there is a strong focus on just AI, not least because you end up solving things with AI that you did not need it for.
There is also something that I think we miss a little bit in Government, although it is talked about a lot, which is the legacy idea. An awful lot of Government’s problems could be meaningfully improved or solved by just sorting out the basics: the data sharing, the ability to communicate across Departments, the good, standard technology underpinning decent HR software, and those sorts of thing. We miss out on those things when we hyper-focus on AI.
Chair: I will move on now, because we are running out of time, but we have a really important procurement issue that I know Martin wants to ask about, so we will take a few more minutes on that, if that is okay.
Q53 Martin Wrigley: Dr Gilbert, you were talking about procurement, which is very important to me, and I think that the Prime Minister has now committed to back British business by default. We have also heard various statements before in the main Chamber about making Britain Britain’s best customer. Do you think that these changes have helped that process, or put it back? If it has changed, what do we need to do? In particular, what do we need to do to bring start-ups in to get contracts—anchor contracts—from Government, which they can then use to build business?
Dom Hallas: This is really important to us. As you would imagine, we are now doing a ton of work on this at the Startup Coalition and are really supportive of the agenda from Government. Why is it important? Because over 80% of current Government tech spend is on 150 companies, and most of them are not British. That is not for want of talented people building technology products in the United Kingdom. We see this as a question of the system and whether it is functioning effectively.
From our point of view, political attention is really important. We have had a very clear message, both from the PM and from the Chancellor, that “buy British” is the way they want to go. The detail of exactly how that policy is structured really matters. For example, paradoxically, the framing of the social value-based procurement that the Government are considering is around job creation. I am all for job creation too, but we have to be clear that more productive firms often create fewer jobs than other firms, ironically, when it comes to technology products. We need to understand what we are really trying to aim for and the impact it could have, certainly on smaller firms that do not necessarily tick all the boxes as effectively, quite candidly.
That political attention is important precisely because the system itself, in its day-to-day functions, does not deliver the outcomes effectively. That is not necessarily at a senior political level or a senior civil servant level; quite often, it is at about people who are making very rational, process-based decisions every day. Nobody gets fired for buying IBM, right? To genuinely change this conversation, to genuinely move it forward, we have to get it to that level and change the incentives that drive that, both in policy and in outcome, and then change the culture. That is extremely difficult; we have to be totally candid about that.
There are some interesting green shoots and examples where we are seeing things being tried effectively. We think that there is much more work you could do on pre-market engagement to understand what is out there in the market before you decide to take yourself down a certain road that often puts you in a position of buying an incumbent supplier by default. It is really important that we do that too.
A lot more could also be done for smaller firms about direct awards of contracts, so that you do not faff around for two years and spend half a million pounds deciding whether to spend £75,000 on something. That happens far too often.
A huge amount could be done in that space. We are super supportive of it. It is just a case of making sure that we can deliver it on the ground.
Q54 Martin Wrigley: I am delighted to hear you quote the phrase “Nobody ever got fired for buying IBM”; I often quote that. Dr Gilbert, on the bigger systems and the bigger companies, how do we get around the default of buying everything from the biggest American company you can find?
Dr Laura Gilbert: Dom has addressed some of the problems here. The Government are doing some quite useful things. NSSIF and the Sovereign AI Fund are looking at early-stage companies. They put direct Government investment in and sometimes own a shareholding. They appear to be aiming to direct that to Government procurement, so you get some sort of guaranteed sell. Of course Government, as a buyer, is the strongest market mover. There is something in enabling smaller businesses because, as Dom said, big tech does not necessarily bring a lot more jobs; what does is more companies proliferating and doing different things. The new technology wave we are seeing enables you to do more and to uncover markets that you have not seen before. Doing it here means that we can lead the world and sell out as well as selling inward. I am happy to see that investment with earlier-stage companies. I would like to see a lot more of it, with a revamp of the procurement rules that makes it easier to purchase.
We are missing one thing. I am going to shoehorn in what I would like the Committee to do. We have a method for telling whether things work, which is an evaluation. In Government, we have a Magenta Book, which tells you how to do those evaluations. We have been running a bunch of machinery of government changes and nobody has ever gone away and evaluated them. Did it do the thing it was supposed to do? What did it cost? What was the impact? We could run that for previous and current machinery of government changes, learn from it and build a structure that says, “Next time we want to make a change, what is the correct or best way to do that that takes into account the cost of the impacts—the recurring costs?”
We should do exactly the same thing in procurement. We have not really run an evaluation that asked, “When we procured in this way, what were the outcomes? What did it cost? How many of them did we end up having to kill? How many did not drive any value? Where did that money go? Did it promote local growth and local jobs? Did it promote national productivity?” and so on. I know those evaluations are difficult to do, but I would like to see that done so that we know how to do it better rather than guessing how to do it better.
Q55 Chair: Before I let you go, I want to raise an important issue that we have not really covered, which is digital infrastructure—our telecommunications networks, data centres and fibre.
Emily Darlington: Satellites.
Chair: Satellites, yes, very good—everything that enables us to operate in the digital sphere and which is so critical to business growth, adoption of AI and all the technologies that run across it. Digital infrastructure has moved from DSIT to DCMS. Do you understand why? Is that the best place for it?
Dom Hallas: Clearly there is some intention to align around where regulators sit so that you have a kind of Ofcom-based discussion when it comes to telecoms and other things. Personally, in the way in which we see the market and the area that we are most focused on—around things like AI infrastructure—I am not sure that that is necessarily the best place for some of those things to work. Based on our conversations, the question will be whether the taskforce, in the highest ambition of the way it is designed, can address some of our questions about how we speed infrastructure development, for example, through its closeness to the top of Government in a way that means that where exactly these things sit within Government does not necessarily matter as much as it otherwise would have.
Dr Laura Gilbert: I think that is right. I think it is about the accountability point. When I first joined Government, Boris Johnson’s Government were rolling out broadband. We built a dashboard in No. 10 with 35 things the Prime Minister cared about, and once a week he looked to see how it was going. That is very motivating. I think I am less worried about exactly where it sits. I am more worried about whether it is funded; who, as a talented individual, is driving it forward; and how accountable the people who are actually delivering it feel. Are they in the spotlight? Do they feel empowered? Can they therefore do their job?
Dom Hallas: I would add a tiny bit to that, which we have not talked about thus far; understandably, given our focus, I have spoken more about the business side. One thing that I think is really challenging about the DCMS structure, as it is currently constituted, is that a bunch of different Ministers have really important jobs within the wider DCMS cohort and then one bit of digital tacked in. That one bit can be the entirety of digital Government and GDS, in the case of Steph, or it can be online safety in the case of Ian. Those things are important in their own right, but you are also doing the arts. You cannot have it be the odd man out. Understanding how to keep that political focus in order to make sure that the work that some very good civil servants will be doing is being consistently driven and supported will really matter.
Chair: That will be part of our role as the Science, Innovation and Technology Committee: to ensure political focus on the issues that matter. Thank you so much for sharing your expertise with us.