Welcome to the Public Accounts Committee on Thursday 18 December 2025. National Savings and Investments is one of the largest savings organisations in the UK: it has some 25 million customers, who collectively have invested over £240 billion. In 2020, NS&I launched its business transformation programme to replace Atos as the single supplier of outsourced operations with a multi-supplier model. In March ’24, the PAC raised concerns about significant problems with procurement for NS&I’s transformation programme, highlighting significant delays and cost increases. NS&I reset the programme in July ’24 and now expects the programme to be completed by March ’28—four years over schedule. Costs are estimated to be £3 billion; that represents an increase of £1.3 billion from 2020. Today’s session will be an opportunity to examine the progress of NS&I’s transformation programme to date. We will also challenge NS&I and HM Treasury, NS&I’s sponsor, on how they are addressing programme changes, and seek assurance that the transformation programme will be delivered successfully to meet the needs of its service users. I thank the chief executive of NS&I for updating this Committee and, in particular, for his letters on 4 August. We like to encourage that culture of openness and keeping the Committee informed, so whatever the outcome of this hearing, we would be grateful if you would continue to keep us closely informed. Thank you all very much for coming. Permanent Secretary, you are extremely welcome. You are a regular attender at this Committee, for which we thank you very much, as you are one of the busiest civil servants in the land. We welcome you all, and please introduce yourselves.

Hello. I am James Bowler, permanent secretary of the Treasury.

I am Dax Harkins, chief executive of NS&I.

I am Matt Smith, chief operating officer for NS&I and senior responsible officer for the transformation programme.

I am Sam Beckett, chief economic adviser and second permanent secretary at the Treasury.

Hello. I am Stephen Farrington, director of fiscal policy at the Treasury.

Am I right in thinking that for four of you—James is obviously a regular attendee at this Committee—this is your first time?

It is my first time.

Sam, have you been before?

I haven’t been recently, but I have been to PAC sittings before.

I have been once before.

Matthew, have you been before?

indicated dissent.

Well, you are all very welcome. Let’s launch straight in and ask Chris Kane to start us off, please.

Thank you. A good Christmas connection there. A key lesson for us from all of that is that we underestimated the complexity of unpicking 25 years of legacy IT with pretty much a single supplier. We did build up our IT capability to deal with that and we brought in external experts to help us with things like system integration, but not of the size, scale and strength needed to deal with the complexity that we uncovered, particularly the complexity of getting the legacy systems to interact with the new capabilities as we gradually phased through transition—the coexistence of the two systems—which we started to understand as we got the different supplier solutions pulled together. Positively, we now have a big, heavyweight system integrator in, giving us the size, strength and scale to push this forward. The NAO Report recognises the difference that that is already making. That integrator is helping us to develop the first iteration of the integrated plan. I think that the key lesson is that we would have benefited significantly from bringing in that big, heavyweight system integrator much earlier in the process to help us plan, formulate and deliver the whole transition. Another key lesson I would pull out is linked to the point you made about capacity and capability, and some of the challenges we have had there. We absolutely understood, as a small, lean organisation going through such a big change, that we had to bring in more capacity and expertise. We have faced challenges as we have gone through that. A good example is that during and post the pandemic, it was a really hot candidate-driven market. It was difficult to recruit people with the skills we needed. We have had more success now, really leaning into the civil service professions and other Departments to get loads of really experienced, really senior individuals, such as our programme director from HMRC, or our commercial director from Cabinet Office, who have helped to improve and strengthen the leadership, giving us a grip. The other lesson that is perhaps a positive one is that we really did prioritise protecting our operational delivery. That meant that, while we have had to work through these significant challenges, we still have been able to deliver. Last year, we increased our net financing from £9 billion to £13 billion, just in this budget, and we are on track to deliver that. We are doing it very cost-effectively. Our value indicator shows that this year we are on track to be £2 billion more cost-effective than raising all our funds through gilts. We are hitting all our service delivery and customer satisfaction metrics. That was an important decision to keep the core operation delivering for Treasury, for our customers and for the nation.

That was an example. We are not saying that that is the cause of all the issues. The key one that frustrates me, as I said, is the underestimation of complexity. We knew that unpicking the legacy would be difficult, but we underestimated how complex it gets with a combination of legacy for some systems and processes, new elements for the new solution on some systems and processes, and new suppliers. That was significantly greater than we accounted for. In hindsight, if we had worked through it in a different way, we could have identified that and brought in a big, heavyweight system integrator early. That would have helped us right from the planning process to configure it differently. It might still have led to the timescales that we needed because of the complexity, might still have led to some of the costs, but it would have been a much more controlled, planned and conscious decision as we worked through it.

NS&I, like a few parts of Government, had this single IT provider that was a bit of a relic of the past. The outsourcing left it unable to reduce costs and to modernise its service for customers, so moving away from that was and is the right thing to do. From a Treasury point of view, I think the NAO Report is perfectly reasonable. For the Treasury the big lesson learned is that we could and should have acted sooner in terms of oversight and support for NS&I—stick and carrot. We will examine that, but the big thing we did was to recommend and fund the system integrator that Mr Harkins talked about. We put two very impressive non-executive directors on the board, with real experience of private sector transformation. We required much closer reporting to Ministers, including of financial information. Also, to deliver on the NAO’s recommendations, we asked David Goldstone—the Committee will know he was most recently in the Office for Value for Money for us—to go in to help us and NS&I to deliver. The short answer to your question, though, is that many of those actions came through 2024 and onwards, and I think that that was probably too late.

I would not necessarily say the assessment was that it would be easier; I think it was the advantages and benefits going forward. One of the key things is that it removed that total dependency on a single supplier, which we currently have in the full contract before we move through. We saw the real risks of that in 2024, when Atos went through significant financial difficulties, and we faced real challenges in trying to unpick that and come up with contingency options to be able to deal with and cater for it. Moving to a multi-supplier model with decoupled architecture allows us to switch out suppliers far more easily, and gives us many more options on that risk. The other area that was key for us was that the multi-supplier model—with decoupled architecture that allowed us to switch out suppliers—drives much more competition. The big, homogeneous single prime contracts, which are difficult to exit, lead to higher costs overall. A third point I would apply there is that it gives a chance to get different expertise from different suppliers to deliver different components.

Do you want to answer that, Matt, as part of the FBC work?

Absolutely. We did rely on advice from external strategic advisers. We also reached into GDS. and spoke with other Government Departments that have been on a similar journey to us—notably, HMRC, which did a big move from a single-supplier model to a multiple model. We made a determination that we should not go from one single contract to 12, 15 or 20 contracts; we went from one to a maximum of five, with the different components. Ultimately, we have landed with some of the providers winning more than one contract, so we are managing three. There absolutely are complexities and different nuances to managing three contracts compared with one. One of the things that we planned for from the beginning is that we would need to develop and evolve our commercial contract management capabilities. We have increased fourfold the size of the commercial team that is managing the contracts in the new model, now that we have these contracts in place, compared with what we had when it was only Atos.

There is always the chance that assumptions are incorrect. You make assumptions based on the information you have at hand, and that is exactly what we did. I used my own experience and that of the broader team, as well as the advice from some of the externals that I mentioned, to make a determination that yes, on balance, it is easier—debatable—to manage one contract rather than three, four or five, but the benefits that we get from that disaggregation, both from a best-in-class service and the resilience in the supply chain outweighed the requirement for us to increase our capability internally. The response or mitigation to the challenge of managing those multiple contracts is to ensure that we built sufficient capacity into the future operating model to take account of that.

This is not just a “finger in the air judgment”—to your point, Mr Kane. The process by which you look at whether you are better off to move changes the business case model, which is the cost-benefit analysis as set out in the Green Book, where you set out the costs and benefits. The costs have risen. It is worth saying that the costs are particularly driven by the fact that the legacy project is costly. The main benefit is that this model looks to reduce the cost, the second benefit is that it is more resilient, and the third benefit is that it allows an improved delivery to customers. NS&I is only now getting digital apps available for its customers, which, as we all know, is pretty vital to choosing our saving options in the 21st century. The legacy contract was struggling to deliver that. Over time, we have looked at the costs and benefits through various processes. As the costs have risen and there have been delays, the decision has become more finely balanced. The full business case from our point of view—because it is being put to us by NS&I—is that it was quite a finely balanced decision as to whether you proceed, stop, pause or do something different. The benefits I set out on the strategic case—that you want to lower costs because you are stuck in this high-cost legacy solution and you want to deliver more resilience and better customer service—are still there. I was only coming in to say that it was not just about taking people’s views; there is a hard and fast set of rules and processes by which you judge these things. I hope that is helpful.

As we progressed through the various business case iterations from the strategic case to the outline business case to the full business case, we developed more of the detail around not only our main central plans, but the alternative plans. We included input from the IPA in its reviews, from the Cabinet Office and from the complex transaction team, taking it through HMT’s TAP process. We did look at those items and built up the information. The problems we incurred in trying to modernise NS&I would not have got better if we remained with a single prime. The issues of total dependency on a single supplier would continue, as would the difficulties of transparency, visibility and control of the entity. It would also continue the risk that we saw come very close to materialising in 2024 when we had the Atos financial issues. We had to spend lots of time, energy and resources developing the next level of detail of the continuity plans, and they were difficult, as a consequence of the model, the construct and the legacy that we had at that point in time.

No. Matt can help me out on this. At the time, that was a recommendation that we would have taken forward in how we progress it.

You can always do more detailed analysis. We assessed what we had in front of us at the time and what we thought was the right thing to do, based on our advice, the strategic outcomes we were seeking to get to and, particularly, wanting to move away from that legacy prime outsourced model. We did manage the risk not of going from one to 12 or 20, but of going from one to a handful of suppliers. That felt like it was a risk that could be mitigated, as I explained. But I agree that there is always more assessment that can be done, and one of the lessons learned is that we could have done more in that space. I am not convinced it would have changed the strategic outcome we were trying to get to in terms of moving away from that one prime outsource model or outsource partner.

Thank you very much, Chris—very good questions to start with. If I could stick with you, Matthew, you told Chris that, before you started on this programme, you quite rightly went around various people—GDS and others—to get their advice on what would be involved here, which is good. They will have told you that big digital transformation programmes are always extremely difficult to achieve. We know from the medical record transformation, going right back to the early 2000s, that these are very difficult things to do. Given that, and given that you, Chris, have talked about a transformation not into one contract but into several different bits—in other words, disaggregated—being even more risky, why did you not consider taking one service at a time away from Atos and dealing with that individually until it worked, and then taking the next one? I hear what you say about the weaknesses of Atos, but you managed to address those. In other words, why not do it sequentially rather than all at once?

That is a really great question. Our original plan was that we would peel services away. We had four transition states. The first was to build the digital integration: the technical backbone. Then we would layer in the customer experience: the mobile app and website. Then we would do the non-digital aspects—contact centre and back-office processing—and finally the all-important banking engine; that is to say the system of record and all the things that underpin premium bonds. That was our phased approach. Unfortunately, we hit an issue as we progressed through the procurement cycle. We ran the procurement exercise for the second phase, the customer experience, but were unable to award a contract because none of the shortlisted bidders provided a compliant bid. We therefore had to rerun that process. That delayed the programme by about a year and therefore concertinaed the time that we had available. We found ourselves in a situation where we had planned to do it in a more sequential way, as you described, but then had to look at the end of the Atos contract, which at that point was March 2024, and see that, because of the delays in that package, we were being concertinaed into a shorter time period. Again, with hindsight—Dax touched on this earlier—we should have taken a step back, re-evaluated and said, “Okay, there’s a date in 2024, but there’s too much to do in that time period, so we need to take stock and make sure we have a viable implementation plan.” That is the key lesson that we have taken away from that.

Okay. Dax, the permanent secretary has laid out very succinctly what you hope to achieve from this whole programme. The question is: will you? What will the reset programme actually achieve and how long will it take, given that the NAO tells us that the programme for banking services is a very high-risk operation? I assume it will not all be complete by 2028. If you could be absolutely candid with us, what is your best estimate at the moment for how long it will take?

That work is still in progress. We have the first iteration of the integrated plan, which was worked up with our system integrator and is supported by all our suppliers. That is a huge step forward, given where we have been, and a key part of the reset. That goes beyond March ’28 currently. However, we have identified options that will pull that forward. A good example—we have talked about the complexity from the coexistence of legacy technology and new components as we phase the transition—is that we are looking at some smart ways of reducing nugatory, throwaway work. We are also looking at where we can prioritise the highest impact items from a customer perspective, such as a fully functional mobile app, and reprioritise some of the less impactful areas, such as back-office systems, that do not have a huge impact to pull the timescale forward and reduce the costs. I am confident that both those areas will improve it. We need to do the work to validate exactly where they take that to. We are aiming to complete that analysis by the end of March 2026. We will be engaging with Treasury all the way through that. We will then look at the governance we need to go through in the second quarter of 2026 to approve all those areas. Treasury have been really clear: one of those options needs to be to work within the budget, which is predominantly driven by the timeframe. We will be assessing the option to do that and we will absolutely be looking at all the risks and all the impacts so that we have a full business perspective to drive that.

I hear all that, but this Committee likes to deal in a fairly hard facts, as Chris Kane made clear. I want to know your best estimate of when this programme will be complete and how much more from today it is going to cost.

I would love to be able to tell you that now, Chair. Unfortunately, we are working through that element. We hoped that the first version of the integrated plan would give us that. I can tell you now that it goes beyond March ’28 and beyond our cost profile. We are working to improve that. We have some real options that will help to do that in a considered and thoughtful way, to get the outcomes that we want, but we have to do the work to validate it. If I give you dates or costs now, there is a high risk that they will be wrong. I would not want to mislead the Committee.

If I were one of your premium bond holders, with £10 in premium bonds, hoping to win the £1 million prize to transform my life—and to be clear, I declare that I do have some premium bonds—I would be absolutely heartbroken to think that my hard-earned money, whether £10 or £20, was going into a programme that has overrun to this extent. It is now £3 billion—another £1.3 billion, as the Report makes clear; paragraph 10 states that “costs are estimated at £3.0 billion (an increase of £1.3 billion from 2020)”. I would hope that my hard-earned money, which I would obviously hope might win a prize, would also be benefiting the Government. Do you have any regrets about this whole programme?

As we talked about before, we have a huge amount of regrets and lessons learned from the amount of complexity, and the fact that we did not approach it in a different way at the beginning to try to handle that, to get the capacity in, change our plans and, as you talked about, phase it in different ways to deal with the complexity that is there from a 25-year IT perspective. I would say that I am glad that you are a premium bond customer, and you have £10—and we had someone with just £11 win that £1 million prize. But I would give you reassurance that, from our customers’ perspective, as we have said before, we have protected the operations, so all the customer service remains meeting all our metrics. We raised more cost-effective financing this year—£13 billion—than last year, so we are still delivering on that. The cost-effectiveness of it—the value indicator—shows that this year we are on track to be £2 billion more cost-effective than raising those funds through gilts. On our core operations, which we prioritised and protected, we are delivering for our customers and for the Treasury. However, I accept that it is frustrating. We need to work through the challenges of the transformation. We need to deliver it to keep those customers, because their needs are changing, and they are demanding more. If we want to keep doing that valuable role for our customers, for the Treasury and for cost-effectiveness, we have to transform and update the business. We are frustrated that it has taken longer to do that. All my people back in the office want the capabilities that we can get, so that we can attract our customers and give them the services they want. I share that frustration, but I want to reassure our customers.

Can I be clear that the cost overruns of this programme do not affect the prizes you might win as a premium bond holder, or indeed your savings rate, just in case there was any confusion?

I did not mean to imply that at all.

I think it was clear that you did not, but I wanted to make it abundantly clear for our listeners.

Thank you for that clarification; I am really grateful. The point I was making is that I would be heartbroken if my £10 or £20 going into premium bonds was being spent on a programme that has overrun by billions. Coming to you, Permanent Secretary, we are about to examine—as you know, and hopefully you are going to appear before us—the Bank of England’s real-time gross settlement transformation programme. That was a really big programme—it was £791 billion a day that it operated on. It was right first time and cost £431 million. How is it that the Bank of England can get it so right, but NS&I can get it so wrong?

I am very pleased that the Bank has delivered on time and to cost, and it would probably be a happier engagement with you on that. I think there are two things. Like you say, the Treasury—Dax is the principal accounting officer of NS&I—holds NS&I to account in terms of value for money and delivery, just as we would hold the Bank of England to account, and indeed any other Department. Our role is to be supportive, collegiate and constructive, but we are also a gatekeeper, as the NAO is the auditor. What you have seen throughout this programme, from a Treasury–NS&I thing, is that we have become increasingly concerned about it, setting increasingly high levels of conditions and interventions due to it, and that continues. I can go through those, but the big thing we have done, which comes out of the skills, capabilities and integration issues, is to insist on some professional input into both NEDs and the integrator. We have now asked David Goldstone to pick up explicitly on the NAO Report, because in reading it, it is clear that we want to do that. To your point about what happens next, that continues. We funded NS&I in full for the costs set out in their full business case. I think Dax has written to you saying, obviously, the cost pressures continue, so we will now engage pretty heavily. Obviously, as you know, we hold the purse strings, particularly if anyone is after additional funding, so we will be pushing quite hard to say, “Well, this is what you said you would do in your full business case. That is what we want to hold you to.” That is the next step. What we probably have not done, which might be useful—this project is in flight, but there has been a fair bit of progress on delivery, which Dax might want to set out, because I don’t want to paint too big a picture. But it is our job to hold to account NS&I, as we would others and we would ourselves. And to your other point, that is where NISTA and the Cabinet Office functions come in. We are supposed to be learning lessons from each other around Whitehall, and of course the job of the Treasury, NISTA and the Cabinet Office is to make sure that we can spread best practice. What we have found, and what this Committee must have seen in a number of guises, is that the decision to totally outsource an organisation’s IT, with quite a lot of the intellectual property, as well as the IT, going out there—you saw this in HMRC and DWP, for example—has proven to be a very expensive issue. These contracts have proved increasingly expensive as the IT that they operate on becomes more and more out of date. You have mentioned it in health and at HMRC. So there are lessons learned in trying to change that. But we are still in this transition with NS&I now, so we will continue to hold them to account, and many of the questions you are asking are questions we are asking.

All that is good stuff, but it took you until 2024 to really start to get a grip on all the stuff—appointing two new NEDs, appointing the chairman of the transformation board, and eventually, in 2025, appointing David Goldstone to look at this whole thing. It must have been clear to you—because you were being asked for more and more money—that this was going wrong from quite an early stage. The IPA rated it red in 2022. So why did you as the Treasury not take some of these quite radical steps much earlier in this contract?

As I said up front, I think that is a fair critique and I accept it. I think we should have. I will say that none of the processes was waving this through with unconditional support. At each stage, we are increasingly saying, “You can go forward, but you must—”, and I might turn to colleagues here as to what those things are. But by 2024, you see us saying, “Okay, we need some radical intervention here.” I think it is a fair critique that that could have happened earlier. It wasn’t the case that we knew everything was going fine. The red rating was not news to us or a surprise to us. The interventions ahead of ’24 were conditional on improvements and progress being made.

As the permanent secretary was saying, in particular, we required increasing frequency and greater clarity of financial reporting from NS&I. There was a difficulty, earlier in the programme, of really understanding what the source of the cost overruns was, because there was a lack of clarity on the funding that was being spent on running the business versus changing the business. We have now moved to that point. As the permanent secretary was saying, it took us a bit too long to get there, and the time for NS&I to respond to those conditions being set was more sluggish than we would have liked, but I think we have got to that point now.

The spending review controls require the lead NED on the transformation committee to assure Ministers that the finances of the programme are under control, so that is a regular check-in point involving Ministers. Also, that committee has to endorse the use of any contingency funding so that NS&I maintains a healthy buffer. So we have used the spending controls through the spending review to try to bring more conditionality on the money, which obviously is one of our key levers, as James was saying.

You have earned another question from me—I thought I had finished, but I haven’t now. You are very good at looking at business cases in the Treasury. This business case was blurred. It included some of the day-to-day running costs within the business case to purchase a new system. If we had blurring of the figures, how could you approve a business case?

At the different approvals processes, we got clarity on the cost of the programme to our satisfaction. The strategic case was strong, and the continuity of service was important. At each of those stages, we have gone through a full Treasury approvals process and had external validation and help from Cabinet Office complex transactions and commercial people. Obviously, in between those processes, I would accept that we have not found it easy to really get that distinct view, but I am pleased to say that the situation has improved a lot over recent months. That transformation committee and its scrutiny has really helped. NS&I has brought a programme accountant on to that committee. They liaise directly with the finance directorate at NS&I, so we now have that separate reporting in place. Again, that lack of clarity was less apparent at the approvals process.

Sorry, Sam, I am going to have to press you on this. You are admitting, as the Report lays out, that there was this blurredness. Okay, in 2024, you started to realise this, but why did you not realise that much earlier? After all, the GIAA was telling you that there were problems with the figures. Why did you not send in real experts to try to get clarity on these figures before you approved the business case?

When we take a business case through the Treasury approvals process, we have lots of external validation to try to help us there. When I say “external”, I mean Cabinet Office and commercial. We have had help from GIAA and have put a lot of scrutiny on that. I feel that at those approvals moments, we had sufficient information to take a strategic decision. We thought we had clarity on the programme. In between those times, we found the reporting harder, but that, as I say, has got a lot better.

As Sam was saying, we put in increasing constraints. In December 2022, when we agreed to the first extension of the Atos contract, as part of that additional funding, we made it a requirement for there to be the development of the integrated end-to-end plan and improved reporting to the Treasury. That was a condition of that extension back in December 2022. It has taken us more time to get those in place than we would have liked, but those conditions were recognised earlier in the process.

It is worth giving a bit of an explanation on how we have built the cost model and where it goes. Part of this is that the transformation programme is for the total NS&I business—it is all of our IT, operations, customer services and products in their entirety. A key element of the benefit is that it then significantly reduces the running cost for all those services. Right from the beginning, we have developed our models and business case on that total picture—all the programme costs, all the legacy run costs and the new contract costs as we come through. When the NAO asked us for the total programme view, we were able to do that because we have looked and tracked the individual components, such as milestone payments for contract deliveries, time and material elements for deliveries, and the cost for programme resources. We aggregated that together to give it to them. We have just had GIAA come in and do an audit of our financial reporting and models. Positively, that has come out with a good moderate rating. It confirmed that we followed the Teal Book standards on reporting of financial elements of this one. However, it was a really useful view to have the total programme costs. We have now built that into our reporting, our transformation committee and our audit committee, and it is providing additional useful insight.

Thank you for that, Dax.

Matthew, we have obviously talked about what has gone wrong in the past, but we clearly need to look forward and think about what we are going to do now—or what you are going to do now. Currently, according to the Report, there is not “a complete and agreed integrated plan that informs the Programme end date”. Can you tell the Committee what you are doing to mitigate the risks, which you set out in your letter, that it is going to go over-budget and take longer than planned?

We are going through a planning exercise at the moment. We do have an integrated plan—as Dax said earlier, that has been built in conjunction with our system integrator and all our suppliers. We have never had that before, so we do have an integrated plan. However, it runs beyond the March 2028 date. We are now working through a process looking at the options for reprioritisation of certain elements of that to establish that integrated plan.

So are the options looking at how to bring that date forward to make sure it is within the 2028 end date?

As far as possible, yes—absolutely. It is obviously a complex area, and some things will bring the data in more than others. That is the analysis that we are going through. We will undertake that work through the first quarter of the next calendar year. We are working very closely with David Goldstone on that; I met with him just yesterday about this actually, so we are getting advice from his expertise, with oversight from the new non-execs on our board.

Are you confident that you will find a solution to bring the date back within the original scope of the 2028 date?

I would not want to prejudge the analysis that we are doing. I am confident that we can improve on the dates that we have in our plan. I think that it will come down to the prioritisation that we need to do and the impact that deprioritising certain things will have on the overall date.

In answer to the opening questions you talked about the challenges with procurement and with recruitment of the right skills. Are you confident that those problems have now been solved, and that there will not be renewed challenges further down the line?

We now have contracts in place with all the providers. Atos was the one that we signed last, not least because of the challenges that it has been going through at a corporate level. We have those contracts in place; however, they refer to the original milestone dates that we contracted to. Therefore, once we have established that integrated plan, we will need to realign the commercial aspects of those contracts. There will be a change in some of the dates, the milestone payments and that sort of thing. The contracts are in place but there will need to be adjustments to them to recognise the delays that we have had in the implementation. On the resourcing side, we are in a much better place. We have around 75 to 80 people who are boots on the ground from the system integrator. They have brought much needed capacity and capability in the deeply technical areas; we were lacking that previously. We have also brought on board a number of other people in the programme delivery space. From a starting point of around 200 individuals, before we commenced the programme, we now have around 350 to 360 individuals. That is not a sustained headcount increase; that is an increase while we go through delivery. The target operating model organisation will be roughly 260 to 270 individuals, to take into account the fact that we need a sustainable system integrator capability, the additional commercial management capability that I mentioned earlier, and some of the customer experience aspects. So, NS&I will be slightly larger than we were when we started this, but we will have much greater capabilities in the supply chain.

As part of that, to build on it, we are also developing strategic workforce management and planning over the next quarter to give us an informed view of how we manage it over the duration of the transformation to get to that end state, so that we develop the capacities and have the knowledge transfer—all the important things that we need to do—in a far more structured way.

Thank you, but to come back to you, Matthew, you mentioned commercial realignment and talked about higher headcount. That all suggests to me increased cost.

It does not necessarily mean that the cost will increase; the headcount costs are built into our current business case profile, so that is covered, including the system integrator cost. That was an additional capability that we added in the most recent business case. On the commercial reset, notwithstanding the reprioritisation that I spoke about, the deliverables will remain largely the same. Because we are delivering over a longer period, some of those costs will certainly increase. However, we have built in a contingency provision—so optimism bias and quantified risk—in line with the Green Book processes.

It has been a planned part of our reset process to have this commercial reset. Once we get the integrated plan we reset the commercials around it. That gives us the real solidity of the milestones and deliverables moving forward.

A further question for Dax: if you have not yet got this accepted and agreed plan—and we know that there is still work to do in the first quarter of next year to figure out how you are going to deliver it in time—how can you be confident that you have planned and costed all of the necessary resources?

It is a good question. Part of the process of agreeing the integrated plan as we go through now will be realigning the resources, timescales and drops. Part of that will be with suppliers in the commercial reset and a part will be with us. We have now ramped up from around 220 to 360 on-the-ground roles supporting us there. We will absolutely need to reconfigure that as we go through—it will not be all the same people and skills at the same time. The strategic workforce management is to help us do that. We are now confident that we have the capacity and skills needed and a route to change, amend and flex them to meet it, and to help us meet that plan once we have agreed it and mapped all the resource requirements against that.

So you are confident that you will have all the resources and budget that you need whatever the final plan ends up looking like.

There are two very different questions there. I am now confident that with the routes, mechanisms and people we have that we will have the capacity and the resources to help us deliver the plans when we finalise them. On the budget piece, it is as we have discussed. With the current plan, the interpretation is that it sits outside our budget profile. We have some options to bring that in that we are working through at the moment. Until we have done all that work, we will not know exactly where that takes us to. The Treasury has been really clear: we need to get an option to work within the budget. We just do not know the make-up of that.

That was going to be my next question to James. There is a fixed budget and yet we are not yet sure of how we are going to deliver the outcome. How do those things reconcile?

What you are seeing here is our financial control of NS&I. We are delivering our financial control in three ways and can be totally transparent about that. We recognise, and therefore we have recommended and funded, the integrator. That is extra money than we had planned for NS&I. Pretty much all your questions come to the need for this end-to-end plan and to integrate and we have done that. However, at the spending review, we have allocated a budget that was based on what was then the full business case. What NS&I is saying is that its current profile is now above that, but we are the Treasury, so it is not the case that you just tell us what something will cost, and we will pay for it, whatever it is. We have said that we will give them that amount of money. What Dax’s letter to the Committee says is that they are currently over that. The third part of the financial control is that we have set up the transformation committee under Melanie Moreland from NatWest as the NED. We have said that you need contingency in this plan, as is best practice, and if you want to draw down that contingency, that transformation committee must approve each of the drawdowns. What you are seeing is the Treasury exerting financial control on NS&I, as we do with most things. And as we do with most things, their budget as specified is what they set out in their full business case, and we have not adjusted that up to do that—so they have some choices to make.

As I understand it, what you are saying is that Treasury financial control is it saying, “This is the budget; you have to deliver within this budget.” But if the plan comes out at higher than that budget, you are then saying that they must stick to the budget, even if that then threatens delivery?

Well, that is just normal cost control and colleagues deliver it day in, day out. At the spending review, we funded what NS&I told us were the costs at the full business case. That is not an unreasonable set of allocations. We funded the integrator and set up the contingency side of things. What I am saying is that the latest letter that Dax has sent to the Committee recognises the conversations we are having too, which are that its current flow would see them costing more and taking longer. Our first step is not to say, “Oh, okay. Send us the bill.” Our first step is to say—

Sorry to interrupt you. Eventually, there is a decision to be made.

Yes.

If NS&I comes back to you in March and April next year and says to you, “We cannot deliver successfully within the agreed budget,” is it then your decision whether to increase the budget or to threaten delivery? That is what I am asking.

We would put those options to Ministers and we would expect those options to include, as NS&I has been saying, an option to bring it within the full business case profile that we funded in full at the spending review. There will be options that NS&I can present to us that will bring the programme within profile. That may include some de-scoping. That may include some delay of back-office functions. That is for NS&I to present to us, and then we will ask Ministers to make a choice on the basis of those options and in the light of the competing priorities at the time.

My final question, Mr Bowler. We are talking about going over budget. What do we think is the percentage over budget that you might be discussing at some future stage?

Others will have a much better case than me. The reason that we are setting this out is because this is all in mid-flight. Given the timing of the Committee, we are being totally transparent with you. I don’t know the answer on the percentage. From our point of view, we will expect to hear from NS&I in the future. That information will come from NS&I. It will also be informed by the work that David Goldstone is doing for us, and by the work the integrator is doing for us. Then, as there has been all along, there will be a set of decisions we need to take. In our normal Treasury function, it cannot be the case that you just carry on paying whatever the cost.

I understand that. Dax, do you have a sense? Might it be material?

As we have said, we already have indicative costs for the first iteration of the integrated platform, but we know we can improve. The options that are being looked at right now will materially improve it. We need to do that work. If I give any guidance, it will be misleading.

That is fair.

Dax, I have been reflecting on your answer to me that you cannot give us a timescale or eventual costs for this programme. You knew for months you were coming before this Committee. What I would like, please, is a very clear letter—I don’t want you to rush this—giving a timetable of when you expect to have that information. Then, when you have it, I would like to know as soon as possible—I am sure the Treasury would too—what the eventual plan is, what it is going to cost and when it is going to be delivered. Can I ask that, please?

Absolutely. That is as expected, given we cannot answer all of your questions fully today.

Thank you. Rupert, I am sorry—I am going to take a break now, because we are halfway through. I know you were expecting to come in. You will of course come straight in after the break. The clock is at 11.03—could we be back here at 11.10? I warn all the witnesses that the microphones will be on, so just be careful what you say. We had one point before the Budget where we were hoping that the permanent secretary was going to give us a leak, but it never happened. Sitting suspended. On resuming—

I am slightly at a loss as to where to start. I have been sitting around this table for a couple of months now. When I listen to things like qualified audits, missing data and lessons learned, I think of the poor taxpayers who are paying a vast amount of tax to fund what is now a very large state part of our economy. I then read the following in the Report: “In 2022-23 and 2023-24, the programme was rated ‘red’…Successful delivery of the project appears to be unachievable. There are major issues with project definition, schedule, budget, quality and/or benefits delivery, which at this stage, do not appear to be manageable or resolvable.” I read of procurement errors, a lack of planning and a lack of expertise, and I really question whether you have the right staff to deliver what is an extremely delayed project, which has cost the taxpayer—directly or indirectly, or however you want to look at it—an awful lot of money. We don’t seem to have yet either a date for the delivery or a final cost. My first question is: have you got the right staff, and are you actually holding your consultants to account in the way that a private sector company would have to do to avoid going bankrupt?

Okay. I will cover that question. I would start answering it by saying that as we have been going through this reset and delivering the integrated plan, we have still been delivering and progressing the transition and transformation. Just this year, we have transferred all our front office and back office operational staff to Sopra Steria. We have launched the first iteration of our new mobile app, which uses our new digital integration layer to connect all the suppliers. It uses our digital experience—

My question was about staff, Dax. Have you got the right staff to deliver what is required? You chose to go a high-risk route, and so far you have failed to deliver it. Matthew, I think you have been involved in this for some time.

Yes.

Are you going to be able to deliver this? Have you got the right staff to deliver it? Are we going to be sitting here in a year’s time having the same conversation? As you probably know, the definition of insanity is doing the same thing over and over and expecting a different result. Are we going to make progress and sort this out, or are we going to be having this conversation when you next come before the Committee?

I can maybe deal with the first part, on the right people and the resourcing capacity. As I said earlier, we knew that we needed to get more people in. We have had challenges in different areas. Now we have the heavyweight system integrator—Matt talked about the numbers before—which has given us real strength and scaleability. We have pulled in some really senior and experienced resource from Government—our programme director and our commercial director—and we are getting David Goldstone to help us on that. We have two new non-executives who have real digital experience, and we are recruiting another non-executive who will come and help us on our board as we go through this. We have moved from about 220 roles to 368 or so.

More cost.

That has been built into our cost. We are now in a better position to deliver this from the resource. We have a strengthened senior team, which helps us. Maybe Matt can talk about the other point about consultants.

We are absolutely managing them. We have fixed outcomes, on which they are paid, so there are milestones. We provide sign-off, and that happens only when there is a deliverable. Importantly, we don’t want the system integrator to come in and do the implementation, and for all the knowledge to walk out the door. Knowledge transfer is a really key focus for us. We are making sure that we have, as I mentioned earlier, people in the going forward organisation to receive that knowledge. We have people working side by side with the consultants taking on that knowledge. We have a good proof point on that with our service management activity. That is a contract that we let to consultants about three years ago.

But Matthew, if you read your report, your staff working with the consultants have not been holding the consultants to account. You have been wasting vast amounts of money on projects that have basically not gone anywhere. [Interruption.]

I will bring you in a second, James. Let Rupert finish with Matthew.

My question is, have you got the right staff now? I am going to ask you a slightly impertinent question, Dax. You clearly have a marketing background, and you speak and present very well. Matthew, you have been involved with this for some time. You two are leading this still, and one of you is from marketing. In my experience, marketing people should never be anywhere near systems implementation. That is something I have learned over my time in the private sector. My question is not only do you have the right staff, but have they got the right leadership?

Maybe I can answer this first. I have a marketing background in the distant past. I have also worked in procurement. I led the last re-tender in 2014. I have worked on implementations of big digital services like childcare services and Help to Save. I have a breadth of experience to bring to the table as well. We recognise that, on something this big and complex, we will not have all the skills, hence why we have tried to bring in experience and resource. David Goldstone is an excellent example. He has lived it—done it—right across the civil service. We are going to take all the benefit from that. We are bringing in non-executives with experience to help to strengthen our governance. We know that we need to bring others in to help us deliver this for Government.

You asked a very reasonable question about the private sector and whether we will be here in a year’s time. I really want to bring out the fact that we have recruited to non-execship Melanie Moreland from NatWest and Luke Jensen from Ocado. They have very real-world private sector experience of big transformation.

Corporate sector experience, I would say, by the sound of it.

Well, not public sector. They have been great. Dax cannot get his contingency funding unless Melanie says it is the case, and that speaks a bit to, “Are they in control of their consultants, or are the consultants in charge of them?” I just wanted to bring that out. We looked at this in the same view that you are, and we looked at who could help them to deliver. Those two individuals have, I think, been doing a really good job to help in that regard.

Okay, thank you for that. I now have a question for Sam. When you answered the Chair’s question about premium bonds, although it is not going to affect the prize money, do you accept the fact that it will affect the amount of money for you to make available to pay for public services? The fact that we have wasted a lot of money on this IT procurement is going to affect the amount of money that you can make available to pay for public services. Is that not true?

You are right to say that those costs are being held by the taxpayer. In that sense, when we think about the overall amount of money that can be spent in the public sector, it will affect either the amount of spending that can be done or the amount of borrowing that we have to do. It has a real cost to the taxpayer.

So it does have a real cost to the British taxpayer.

True, but as an institution, the NS&I’s sole objective is to help us finance the borrowing that we do, and the aim of this project is to reduce the cost of the way that we do that.

As you know, James, it is basically a state monopoly. That is what NS&I is. We know that.

We can talk about that, but I take your point.

We can debate that on another occasion. My final question is this. You have David Goldstone, whom you have referred to a lot. I saw from the report that you brought him in. I personally would like to see this Committee have a note from David Goldstone. I am representing the taxpayer, so I would like to know from him, with his huge experience of British public sector delivery—including the Olympics, I gather—a realistic estimated cost and timetable for complete delivery of this project. I think it has got to be made available now, and there have got to be real consequences if there is further slippage with what I already think has been an unacceptable splurge of taxpayer’s money that was easily avoidable.

I think NS&I rightly need to write to you setting out what the cost and time period is. I am sure we could ask David to validate and add to that piece of work.

David is an integral part of the work we are doing, which, to reiterate, is the integrated plan. We are working to do exactly that; we will have the plans and the costs, and that will inform the decision. As you asked for, we can give you an update as soon as we have that, to set it all out so you have real clarity. That will complete the reset and give us more consistency and more deliverability, which we are all striving for.

Try and remember that every pound you waste means that somebody else is going to be taxed on their business or their farm, and there are going to be real consequences for Britain if the public sector continues to waste money on the scale that, it is quite apparent to me, is happening.

Permanent Secretary, you have already been candid enough to say that you could have been more proactive in recognising the problems. Presumably there is a bit of a deadline on this, because you are going to need to plan, in line with normal Treasury procedure, how much money you are going to give these people for their transformation programme in the next financial year. You are presumably going to need to have that information well before the financial year starts, are you not?

Mr Farrington might be able to give a more a detailed answer. We have allocated money for three years to NS&I as part of the July 2025 spending review. They have a budget for the next three years, as all Departments do. The question they are looking towards is whether they can deliver within that budget.

What we have done for this year is also introduce ministerial review points—that was a condition of the spending review settlement in the summer. Those are quarterly review points at which progress on the programme goes to Ministers, and we seek approval from Melanie Moreland, the non-exec who is chair of the transformation committee, on whether NS&I are making sufficient progress that the funds allocated for the programme can be released. We are currently concluding the first review point, and she has advised me that those funds can be released, subject to further conditions on NS&I. We have increased our in-year monitoring and release of funds to NS&I.

Permanent Secretary, could the Committee see those quarterly review point reports?

I will certainly take that away.

You take that away, and if necessary, we would have it under normal reading room rules, which you know that the Committee is capable of doing, and it would not leak. But we would rather have it on an open basis. That would be really helpful, so that we can be satisfied of where the trend of this is happening, and whether it is within that three-year budget.

The key thing that is going to determine that is when they come back with the integrated plan, and the costs and times of that.

I read that as if it was talking generally across programmes. And we had had a culture review by GIAA, which we shared with the NAO, and they referenced some of the points that we know we need to address as we work through the process. On good news culture in particular, I would say, if I look at our organisation, that people are really committed to NS&I and to its purpose, to its values and to the transformation. On occasion, that can lead to a really “can-do” attitude, which is invaluable. However, there have been occasions when it has led to people being over-optimistic in delivery. What we have done is to bring in really experienced resource, such as our programme director, and the likes of David Goldstone. They will particularly give check and challenge as we develop this integrated plan and reset the programme to help us through that. As for culture more widely, we know that it is really central to delivering a transformation. We got some clear output from the culture review that we did. We have built that—the factors in the findings from there—into things like a leadership development programme that we are running now. We have a new management development programme that runs in parallel to that. I am really pleased to say that we have just got—probably in the last week or so—the results from the civil service people survey. Our engagement score has increased from 66% to 69%; willingness to challenge has increased from 65 to 70%; and confidence in leadership and management has increased from 61% to 67%. That is way above civil service averages, which is positive. We will always have more to do on this when we are doing a programme of this nature. So we will keep investing and keep driving that forward.

It builds on the first part of that answer, I think. I do think we have seen people being over-optimistic in some of the deliveries previously. I think that some of those experienced resources we have brought in have been there and done it before, and will provide that check and challenge. Just recently we have talked about how we will get a review in to look at the integrated plan once we have got that, to give us some independent check and challenge. So we are trying to lean on the tools and connections that we have across other Government Departments and across organisations such as NISTA that will help us, because we absolutely understand that as we get this integrated plan that is agreed by our suppliers, we want that real, consistent delivery against it. So we will build all of that in.

You make a good point. The tone from the top is vital; I absolutely understand and live by that. As we go through this, I am really mindful of checking and challenging, and pushing, and giving us the contingency and the certainty of those deliverables, as we go through this process. Even as things come into ExCo, as they come into our audit committees, things like ensuring that we have had that right assurance, and using people like our risk department to give us independent validation of that, are all important. They are some of those smaller day-to-day steps that you are referring to.

100%—I absolutely understand that.

That is a good question; you have given me some pause for thought. What would be an observation that I would take from the leadership development programme? It was good to get that cohort of leaders together. It was really interesting that when we raised issues—we have challenges that we have outlined today in the transformation programme—there was not always the recognition that those leaders in the room were the people who were going to address those issues and drive it forward. It is not just me, or just my ExCo. It is not the experts we brought in. It is those leaders in the room that have a key role. We are working on that. People have taken that away as a recognition. As we have gone through the different phases, we have seen that begin to develop and mature. But there is absolutely more work to do on that. A key part of the transformation is that we know we need to develop the capabilities of our people. Moving to multi-source, where we insource key components and we are now responsible, takes a shift in mindset and accountability. So the strategic workforce planning and management is a key part of how we continue to develop that capability to get the change.

It is another good point. We have worked really hard. We have been going through these challenges for a number of years now. It is taking a heavy toll on all of our people. It is difficult for everybody to work through. So we have invested a lot of time in working with them and communicating with them, particularly on some of the achievements we have had this year. We have moved forward a lot. We have successfully done the staff transfer of the front and back office operational staff to Sopra Steria. We have successfully done the first iteration of the mobile app, which really takes us forward to future technology. Atos has successfully done the first seven milestones of the critical core banking transformation. You cannot overestimate the power of getting this big heavyweight system integrator in. We have really focused on those. We have been honest about the challenges that still remain and the fact that we need to work through this integrated plan, and the fact that we know we need to strive to work within the budget and work with them on that. I think the hard work around that has helped us. They do understand the challenges we still have, but having them engaged and working to solve the problem with us is invaluable for us. It has been a key part of our strategy over this past year.

Yes, we definitely recognise that. We know we need to develop the capabilities of our people for the new model. We also know that means taking more accountability and more responsibility, which is a change from the role that we have got now. We are doing a capability review as part of the strategic workforce management to try to help us with that. That is a big task that will take time to work through. It will take time to develop it. We are clear we have got elements to do. It is not all good news culture, but it is a positive that engagement has increased. And the cultural review was over a year or so ago now, and our engagement score did dip last year because of the challenges we have had. I think this is a consequence of some of the work we have done this year in improving it.

I will come in on that. The NAO Report recognises that our governance has evolved throughout the lifetime of the programme, as we have had different stages. We are in a much stronger position now than we have been at probably any point in the past. The permanent secretary has already mentioned the two new non-execs with relevant experience coming in, as well as David Goldstone. We have another non-exec who is due to be recruited in 2026. We definitely welcome all of that. The transformation committee that has been referenced, which is chaired by one of our NEDs, has played a really key role over the first 12 months of its being in place. The insight and outcomes that those committee conversations are driving, as well as being held to account, are driving a new level of focus on things such as the financial management, as well as delivery. I would also go back to Dax’s earlier point. If you look at our business case from 2024, we have delivered in 2025 all the things that we said we would deliver in staff transfer and mobile MVP, and we have onboarded the system integrator. The governance is working from that perspective; it is driving delivery. It links back to the earlier point about staff engagement and staff morale. They are seeing that we are making progress and delivering. This year has been the first year in the lifetime of the programme where we have actually seen, if you will, the rubber hit the road, or a change in the legacy model. We now have three parties, or suppliers, on the pitch, rather than just one. There have been real changes in governance even in the last 12 months, so hopefully that goes some way to answering the question.

Yes, that is something that we need to continue to look at. Obviously, we have just added in the board transformation committee, so that is an additional layer of governance. We need to make sure that we have the level of scrutiny that is clearly required, but also that we are not slowing down the decisions and we are making them in a timely manner. One of the things that we are doing, as a result of a recommendation from the NAO study, is re-looking at our governance to make sure that it is not only giving the required output but allowing us to deliver and move at pace.

I think we have made some significant changes over the course of 2025, particularly in the last six to eight months. We mentioned earlier that we have additional finance capabilities; we have increased the size of the finance team by over 20%.

I believe that is an additional six people.

In programme finance.

Yes, in programme finance. That obviously give us a huge amount more capacity to do the level of reporting and scrutiny of the costs that are coming through in the programme. We have recently gone through a process with both the transformation committee and the board audit and risk committee with the new reporting pack, and they are satisfied that they are seeing the required improvements. It is not yet the finished article; there is still more work to do. I think Dax mentioned earlier that the Government Internal Audit Agency report that we have had in recent weeks gives us a moderate level of assurance on the financial processes. It is a marked and significant change from where we were this time last year.

Yes. We have got the access to the information. I think we still have some more work to do in terms of how we how we present that information. Our systems, like all our IT infrastructure, are quite aged. It is not always the easiest thing to get information out in the form that you would like, but we have the new capabilities that we have brought in with additional people within finance, and using new tools and techniques. The systems integrator is helping us with that as well in terms of enhancing our programme management office. That will give us the level of information we need. Yes, the data exists to allow us to make the right decisions at the right point.

We are working through that as we build and refine the integrated plan, as we discussed earlier. We are going to be working through the first quarter of the next calendar year. Finance is working hand in glove with us on that, alongside my programme team. We are taking advice from David Goldstone as well in terms of best practice there, and we have got regular touchpoints with colleagues from the Treasury, so we are in lockstep as we go through that process.

Matt and the team are working with suppliers on the integrated plan. As part of that, we will get that commercial reset view, which gives us more consistency and certainty around those figures. They are actual figures from the suppliers, rather than our assumptions about where we think we will be. All that gives us more certainty as we progress through this next period.

We are basically doing interim milestones. Whereas the original contracts had fairly comprehensive milestones, we are doing it in smaller steps, so that they are delivering a set of artefacts or a set of outcomes. While we are going through this planning exercise, we are continuing to work on the things that it makes sense to continue to work on. Good examples of that are the staff transfer activity that we have undertaken and the building of the mobile app that we launched a few weeks ago. Throughout those, we have had deliverables at certain points in time, and we have paid them based on those deliverables.

Yes. A good example of that is Atos, not as our incumbent supplier for legacy but as the supplier that is building the new banking engine. Just last Friday, we completed the seventh milestone out of 21 over a three-year period. That is seven milestones in 2025 delivered on time and on budget. That is working extremely well.

To follow up on the cultural point, you said that there has been an employer survey. The organisational numbers are in figure 1, with 254 people. You have then indicated that you have 350 contractors. How many people internally have been asked about the culture in relation to the change programme, and have you also asked all the external providers as well? Otherwise, you are giving us only a partial picture of how individuals could respond. More broadly, from the perspective of the organisation’s culture, how are you measuring the fact that you have a large headcount that is not within the organisation but is still servicing the organisation?

The survey that we have just done is the civil service people survey, which is annually run and co-ordinated centrally. Only civil servants take that survey, so the scores that we mentioned earlier, including the 69% engagement score, refer to the civil servants within NS&I. We take the temperature regularly with the consultants who we have working with us on the programme. They are working with us side by side. We have regular review sessions with those consultants. We have a monthly touchpoint, for example, with the systems integrator where we will talk about the progress over the previous month and about any concerns that we might have either on delivery, the fit of individuals, how people are working together—ways of working, if you will—as well as how we might need to adjust anything for the forward-looking period and the work that we have planned over the subsequent month.

If I may follow up on that, let’s say there are 20 people in house who are dealing with the programme change and hundreds who are not in house, you are basing your judgment of the culture on just those 20. Is there a risk that you are getting a particular view about the programme change from a small number of people, but that view might not be realistic because there are so many other people working on it who you are not asking?

The proportionality is important. The 250 civil servants who took the survey are part of the entirety of the 350 or 360 people who we have working within NS&I. That is not 360 consultants; there are roughly 70 or 80—something like that—and that figure changes based on where we are in the programme. Roughly 25% are consultants, so the lion’s share are civil servants and that is represented in the engagement score that we see. Hopefully that is helpful clarification.

I might come back to that in my questions.

You have further questions to ask, so please continue.

Okay. I have a series of questions specifically on the success of the future programme, including this element on recruitment. Your current delivery metrics show strong performance in net financing, customer satisfaction and value to taxpayers. Do you think that the metrics you currently have are a fair way of measuring the entire organisation, or should they be expanded to include capability and programme risk?

I understand the point you make. We are assessing the progress and the risk around the programme through all the programme governance and the elements we go through there. There absolutely is focus and scrutiny on the progress that we make. We are doing that. That does include the risk elements. As part of this reset, we have reinforced our risk process. We have set up a specific group to review risk with our assistant risk director, our programme director and others, who meet weekly to review both strategic and operational risk and give our governance a statement of their assessment of risk. That goes through our executive risk committee and our audit committee. I understand your point; I think that issue is dealt with separately to our service delivery measures, which are on the long-term enduring elements of the reason that we are here—to serve customers and raise net financing.

I asked that because paragraph 20 of the Report indicates that there is “more work to do” on governance arrangements and risk management processes. It recognises that “arrangements for managing Programme risks and issues were inadequate” and that that has a potential knock-on effect: “an extremely high level of risk because of its potential impact on customer data”. In paragraph 18, it says, “Stopping these services without sound planning would cause significant disruption for departments”. The reason I raise the risk element—I will come on to it in more detail—is that, in my view, it should be part of your key delivery metrics because if you get it wrong, there are very real business impacts that are highlighted in that Report. I ask again, on your risk management profile, do you think that there is a capability problem within the organisation that has been highlighted? Do you think, because of those issues potentially impacting future business, that should be part of a key delivery metric that we measure you on going forward?

To unpack that a bit, we have strengthened our risk directorate’s involvement in managing that. The risks were managed within the programme at one point, but we have taken them out and got the risk director to do it, in a combined effort—that is really helping us. We accept that we do not have all the expertise in these areas so we use GIAA to help us, but we also go to external assurance specialists. We have used a number of them for our staff transfers; they assessed our first mobile app launch. We will absolutely get them to come in and talk about the big data transfers of the complex banking engine transformation that we are focused on, to help us as we go through it. On the other part of your question, about the metrics, I understand your point exactly: this could impact the business. One of the main risks we track is the risk of business interruption as we do these migrations. We are absolutely focused on all the mitigants to do that, so you are right; we absolutely need to focus on it. Whether that then becomes our service delivery metric is a different point. I am sure that all our governance, our board, our audit and risk committee and my Treasury colleagues are absolutely focused on those risks. I am unsure whether a service delivery measure would add anything to that.

The reason I suggest it might is because, clearly, this programme—as my colleagues have suggested—is not only delayed; it is over-budget. You are right that you are hitting your current key delivery metrics—I know you keep going back to that; as you are a monopoly service, it is good that you are—but the concern is that you have not been hitting the metric on the programme change element. One would suggest that, if you have not been hitting it, that should be potentially a key metric that we measure you on as part of your key delivery going forward. I am just saying that as a further observation from the public. If you have hit all of the other key objectives except one glaring one, it would surely be logical that we measure you on that going forward as a key delivery metric.

Matt, the organisation and I are absolutely measured on this. Most of our time, energy and focus in governance has been on the transformation and the issues there. I think there is absolutely that focus. It is a key part of our report and accounts. This is a transformation of our whole business, so we are absolutely focused on all those risks as we go through it. I want to reassure you that there is no walking by this. There is an acceptance of the challenges that we have had and the need to do more. It does impact the perception of our performance.

Okay. I will move on to contingency around supplier risks in particular. Looking at Atos, and financial and other organisations—you are moving to multiple platforms—is supplier risk, or mitigation of supplier risk, a concern for you?

It is, absolutely. For any organisation that has a heavily outsourced model—as we do and as many other Government Departments and private sector organisations do—supply chain risk has to be a key area of focus. We were monitoring the situation with Atos very early on in the cycle in 2023, and followed that throughout the process, working with colleagues not only internally but in the Cabinet Office. Obviously Atos is a supplier to broader Government, as are Sopra Steria and IBM, which are the most recent entrants into our supply chain. As part of the enhanced commercial capability that I spoke about earlier, we have supply chain risk as one of the key capabilities in that. That is something that we will keep a sharp focus on as we go forward. Some of the names I have mentioned are large organisations, but history tells us that we need to make sure that we are focused on that and ensure that we have appropriate contingency plans. A big part of this is working with colleagues in the Cabinet Office and the marketing strategic suppliers team to make sure that we have the insight that they have on the broader opportunities and risks related to the suppliers that we are working with.

On risk, it says in paragraph 20 that you have introduced a new risk management framework. With the issues of supplier risk that we have just mentioned and the programme risks that I started with, do you feel that the risk management framework that you have introduced is cogent enough to pick up on some of those elements and ensure that, for instance, we do not have potential impacts on customer data, or failures that might lead to user Departments to consider options for service delivery somewhere else, as the Report highlights?

Absolutely. The risk management framework is comprehensive. It is a set of principles that can be applied both to strategic risk and operational risk, and we are reviewing both those elements on a monthly basis in our executive risk committee and on a quarterly basis with the board’s audit and risk committee to make sure that we have that clear view across the whole estate.

Given the risks that scream from the Report, it might be helpful for us to see the new risk management framework that you are putting in place. Clearly, there is possibly a future ongoing liability if things are not correct on that. Ensuring that we have those robust risk management processes would be helpful.

Right.

If I can move on, recruitment is obviously a key element of the success of the future programme. You have an internal capacity, which you have increased. I notice from figure 1 that the staffing numbers have increased. You are using a lot of external contractors. You are also borrowing staff from Treasury and elsewhere in Government—I think I heard that earlier. You have had to bring in a lot of people. Are there recruitment challenges for you here? Or is it that you just do not have the budget to bring them in internally, so you have to use external contractors?

Previously we did have recruitment challenges. In the post-pandemic period of the so-called great resignation, we were really challenged, particularly in the technical space, with the salary levels that we can offer and competing with private sector organisations. That is particularly in things like data architecture and technical integration-type roles. We have come through that process. We mitigated that by working closely with partners in industry, using some initiatives that were on offer through the Cabinet Office to take on loans from industry, so we have essentially had people on secondment into NS&I. Those secondees have worked side by side with my architecture team and transferred that knowledge. At the same time, we have been recruiting into the vacant positions, so we have mitigated that to a large extent. It is natural with any kind of programme that it is a temporary organisation. You would not want to do permanent recruitment, so it is quite usual to have contractors and loans and so on. The benefit we have seen from taking civil service colleagues on loan from other Departments is that they bring their experience with them and hopefully learn from the process we are going through and take that back into their home Department and the broader Government. There is quite a neat ecosystem there that is working quite well for us.

Lastly, on milestones, Chris mentioned the milestones that you are holding contractors to account by. Paragraphs 3.7 to 3.9 in the Report note that there have been previous reviews—the Gartner review, for instance. You told the NAO that you “had resolved 23 of the 42 highest-priority end-to-end design issues”. That review was from 2020. There are still a number that are outstanding—that is milestones that have not been reached, presumably within a timeframe. Are you able to audit the milestones robustly internally to ensure that you are getting that? Do you have the management, the headcount, the budget and the capability to audit that robustly to ensure the milestones are met by your external contractors?

Absolutely. As part of the processes we have put in place—I mentioned earlier enhancing our commercial structure—we have contract managers in there, who work with the business owners to ensure that we have complete clarity in terms of what the expectations are from the supplier, and that they are meeting that expectation. It is referenced in the NAO Report that we were in the process of putting together a contract management framework. We have finalised that now and we are in the process of rolling that out across the organisation. At the moment, roughly 25% of our colleagues have been through that training. That will be ongoing over the next few months, so that everyone is acutely aware of what the expectations are and how they are expected to manage anything that goes off track.

You have spent £43 million on external consultancy contracts. I have probably not read many Reports where there are so many different consultancies mentioned. There is Capgemini throughout—there seem to be a lot of external contractors highlighted, which then begs the question of whether they have been measured in the past correctly. Is there an internal organisational problem because you are having to go out all the time to try to identify those issues, and is that justifiable for the taxpayer? Do we think that constantly going out to so many external contractors is sustainable and justifiable? That is a question for Treasury as well. Is that not something you have fed back to NS&I? At paragraph 3.8 of the Report, it says you appointed Capgemini, another contractor, in May 2025. They are going to help you integrate. Have you had that report back? Have they identified any issues with systems integration that we need to be aware of?

They are the systems integrator—or the systems integration partner. As we mentioned earlier in the session, they worked with us and with the suppliers to build an integrated plan, so we do have that plan. It does not meet the data and time profile that is required, and hence we are doing the work in the first quarter of next year to refine that and find a suitable landing position. In terms of our reliance on consultancy, it goes back to NS&I being a very small organisation. Even though we are probably a big brand, we were, at the start of the process, still an organisation of 220 civil servants with a heavily outsourced model. Undertaking a transformation of this nature required us to bring in external expertise to understand how we articulate our designs and our requirements for procurement, and undertake the procurement activity. It would not have made sense to recruit those people in to be part of the civil service, so we brought in that external expertise. We have not had all these consultancies on the pitch throughout. There have been certain intervention points. Legal counsel helped us with procurement, for example. We had another agency helping us with enterprise architecture capability to build that capability. It is all part of our evolution from essentially providing oversight of the delivery that Atos was doing on our behalf, to being that intelligent client and having that ownership and accountability for the overall business model.

On the last few questions, it is probably worth saying that, from our point of view, the NS&I has its own principal accounting officer and its own board; we have to judge how we engage. We operate an earned autonomy model. If everything is going well, we will leave it more alone, and if not, we will intervene more. We are at the quite interventionist end of that spectrum. I will not repeat it, but we have listed all the various things we have required, for the reasons that colleagues have set out. There is a need here to get past all the external advice, although it is helpful and needed, to get to a status quo where the body can operate its new multi-contract thing strongly and without the need for lots of people pushing in on that. That is where we are headed. You are seeing this programme in flight, although it has been in flight for a while. From our point of view, we are at the interventionist end of how we do it. We are not doing that to the Bank of England’s model; it will have set it up differently. That is how we operate.

I appreciate that clarity. You have admitted that you are at the interventionist end. Do you think that the organisation accepts that it is at the interventionist end? Does the public understand that, given we are where we are, there is a systemic issue that needs to be identified within the risk profile of the organisation?

I think it does. I do not think what I am saying is news to anyone. The point is that this was a complex and difficult project for an organisation to go through. There was optimism at the start, and you talked about culture there. The reality is that it required hard-nosed interventions. The integrator, Capgemini, has been very necessary. You break something up into different parts and you have to bring it end to end. Those are sensible things. I am talking about our separate roles and how we do it. I would say that it has always been constructive. That is how we operate. But, like you say, there is a lot of hindsight. If it had all gone absolutely to plan, we would not be saying, “Here’s two new non-executive directors; here’s David Goldstone; here's Capgemini.” But I think it was the right thing to do.

I think we are in a good place. The evidence we have given you today has been about whether we should have got to that a bit earlier. But I think we now have the structures in place: the governance, the oversight, the help, the integrator. Our answer to reading the NAO’s Report is, “Let’s get someone in to make sure that the recommendations are delivered.” I think we are in the right place. You are seeing us in flight, hence the answer to the questions—exactly when is understandably still being worked through. I think we are in the right place. Quarterly reporting to a Minister on exactly where things stand is at the interventionist end of the spectrum.

I would add that the interventions we have been talking about are specifically targeted at the problems identified in the NAO Report. The four things it lists as the root cause of the problem are the lack of a systems integrator, the lack of an end-to-end plan, insufficient expertise and insufficient contract management. All our interventions have been specifically targeted at precisely those issues. That gives us some more confidence for the programme going forward.

I have a few catching-up questions. First, the Report mentions the other Department contracts that you have—banking contracts, and so on—that you no longer wish to carry on. Has there been proper discussion with those Departments so that there will be a proper handover plan and they will not be involved in unnecessary cost in picking these services up?

Absolutely. We have been working with the Departments for these services over the previous months. We are working up options analysis right now that we will take to Ministers in March next year. In all of this, I want to reassure everybody that we are working collaboratively with those Departments for the best outcome for Government. We are absolutely focused on ensuring that there is no disruption to any of these services as we work through them.

Thank you for that answer. Matt, going back to your answer to Tristan Osborne about systems integrators and Capgemini, I must say loud and clear on behalf of the Committee that so many of these large Government projects have gone wrong because of the systems integration at the end. They think they are almost there, but they are not. It does not matter whether it is HS2, the Elizabeth line or some of the more complicated military projects, the systems integration is always the problem. Alarm bells began to ring when you said in answer to Tristan Osborne that you had had a report from Capgemini about systems integration, “but it did not meet our timetable”, as I think your words were. What does that mean?

That is the integrated plan that we spoke about earlier in the session. Capgemini came in as our system integration partner and worked with us and with the suppliers to pull the different component parts together. We now have an integrated plan, but it does not meet the timelines that we had in our full business case; hence the work we will be doing between now and the end of the first quarter of the next calendar year to refine that and bring it back as far as we can.

The alarm bells are still ringing. Can you put a bit of flesh on that? How far out of kilter do you think we are?

It is difficult to define that. There are elements of reprioritisation that we are going through. For example, one of the things that we had originally planned to do was build all the new capabilities to the legacy—or basically build the new website and mobile app on to the legacy IT. One of the initiatives that we are looking at to bring that plan forward is not to incur that nugatory cost and time of building to the legacy, but to build to the new capabilities of the new banking engine first. We think that it is quite a significant opportunity to bring in the plan closer to the timeline that we had in our business case. There are a number of different aspects that we are looking at.

Can I stop you there? With this initial report that you have had, are we talking about months or years beyond the original business case?

Certainly not years. It is a number of months, and that is something that we need to work to remediate.

I can maybe add some reassurance: it is a number of months. The system integrator and the suppliers advocate these options to pull them in, so they are currently working through them. The nub of what you are trying to say is that we should not ignore the system integrators’ advice and strive for something we cannot achieve. I guarantee you that we are not going to do that. We are going to take all the options to do this. We will look to improve it, because we know there are financial pressures, but, above all, it must be deliverable, and we must have confidence that we can go through this. We will work through that and look at the risks, impacts, consequences and deliverability as part of all this analysis that we are going to do by the end of Q1 next year.

I am grateful for that reassurance, Dax. The Report makes it quite clear that you regard the end of Atos’s contract as being 2028, but then you are talking about other bits in the Report that say it might be renewed beyond that date. Bearing in mind Atos’s deteriorating financial condition, which is the whole reason why this whole thing started in the first place, what discussions are you having with Atos to renew its contract beyond 2028? Based on everything that we have heard today, it is not going to be completed by 2028.

The agreement that we have put in place with Atos runs until 2028. That was in line with the approval that we received of the FBC from Ministers last December. However, it gives an option for an additional one-year extension, but we would need to take approvals before we can exercise that option. There is a further period beyond that for transition assistance. To clarify, Atos was in financial distress during 2023-24. It worked with its banks to put in place a finance restructuring deal in December 2024, and we see that all the metrics are improving on that. Again, we are working very closely with the Cabinet Office, which is monitoring this more broadly across Government. I think Atos is coming out the other end of its financial issues. At the moment, it is delivering both the operational system absolutely to our system levels—hence we are delivering for NS&I customers—and the build of the new banking engine, as I mentioned earlier, with seven out of 21 milestones on time and on budget. We are very happy with where it is. However, extending Atos beyond the current March 2028 date is something that we will need to look at as part of the planning that we do in the first quarter.

Can we be assured from both of you—you are the SRO, Matthew—that there will not be a gap? You are saying that, if these other contracts are extended beyond 2028, Atos will continue to operate for as long as it takes until you are able to migrate into the new programme.

Absolutely. Service continuity is the paramount priority for us, so there is no question. That is why we put those contingencies in place, so that, should it be needed, we have time in the contract out to 2031.

This was not quite clear to me. Tristan referred to the NAO Report, which states, “the core banking engine carries an extremely high level of risk”. In my 12 years on this Committee, I have never heard the NAO go in with such strong language, yet you say that Atos is going to perform that programme. Is it going to perform it temporarily or permanently? Are you confident that it will be able to deliver it?

It is building the new banking engine with a third party that provides the banking engine software. They are the two companies—Atos and the subcontractor—that provide the capability in the legacy estate today. We are transitioning from that legacy banking engine to a new one, and that is what Atos are working with their subcontractor to build. We are very comfortable with the approach that they are taking, but we are ensuring that we undertake assurance throughout. Data migration, obviously, is a huge consideration, and that is something that we will be looking at in minute detail.

I am still not quite clear from your answer; is Atos’s role in this banking engine a transitional one, or are you expecting them to run it permanently?

They are building the new banking engine, and they will run it until March 2028, unless we decide to exercise a contract extension. I think there is reference in the NAO Report as well that, because the Atos contract currently runs to March 2028. We always planned that we would need to go back out to market, not for a new banking engine, because they will have already built that, but for someone to run it, essentially, and make sure that the maintenance is there.

Can I come on to the figures now? I don’t mind which one of you answers this, but can either of you tell me what the exact spend on this programme is, to date?

The overall programme cost is around £800 million. That is the building of the assets, not the running. We had sunk costs to March 2024 of around £111 million. Over the course of the subsequent almost two years, there is around another £250 million, I believe—I would need to have a look and come back to you for the exact figures—of both build costs and elements of run costs as well.

I am struggling slightly because there are two different figures in the Report. Paragraph 10 of the summary says, “NS&I estimates the Programme delivery cost to be £841 million”. However, paragraph 1.8 says that your next three-year budget is £983 million. Which of those figures are you working to?

The £841 million is the entire build cost; that is building the new systems and the new capabilities. The budget of around £980 million includes the run elements of that for that three-year period. So the £980 million has part of that £840 million, but also includes the run cost for that three-year period.

Okay. I would appreciate a candid answer to this: the Treasury has set you that three-year budget; do you think you will be able to live within that three-year budget?

That is the work that we are doing now; that is the planning work that we are going through. It depends on a number of factors, but we absolutely understand the direction of travel and we need to have an option that fits within the budget. But I do not want to prejudge the work that lots of far more experienced people in financial analysis are doing. That is the work that we are undertaking over the next three months or so.

All right. We will end the session there. I will just say to you that there are still lots of unknowns, not least of which—you were not able to answer this today, understandably I think—is how much this programme is going to cost in total and how long it is going to take. This Committee is hugely interested in the answers to those questions, so, as I said earlier, Dax, please keep us closely informed. We look forward to having those Treasury quarterly reassessments, if we can; those will give us an update on exactly what is going on. I thank all of you for your time today; you are all busy people, particularly the permanent secretary. We are grateful for your clarity and your presence today. The uncorrected transcript of this hearing will be available in the next few days. We will thereafter produce a report with, I am certain, several recommendations, which no doubt you will be keen to look at and see whether you can accept—hopefully you will. We hope that this programme, now that you have all this expertise in place, will run smoothly from now on and will be delivered on time. Thank you very much.

Chairman, could we have the report from David Goldstone—or, as I have nicknamed him, the Ghostbuster? Could we see that? I think it is important that we see that.

Rupert, I have not had advice from the Clerk, but he is not here today. He is separate. Unless you ask him, or we could correspond with—

He has gone in to help. I think that NS&I should write to you with the answers to your questions. I cannot speak on behalf of David, but I am sure he could give his take, validation, and thoughts on that.

It does seem to be an integral part of delivery, though.

Absolutely. We are not trying to—I think the answer is a version of yes.

We will correspond with Mr Goldstone. He has always been very helpful to this Committee, and I am sure he will do what he can.