Official Report: Minutes of Evidence

The Chairperson (Mr McCrossan): From the Department of Finance, I welcome Neil Gibson, the permanent secretary and accounting officer, and Paul Duffy, deputy secretary. We are joined by Stuart Stevenson, the Treasury Officer of Accounts in the Department of Finance and Dorinnia Carville, the Comptroller and Auditor General, along with Rodney Allen, the chief operating officer in the Audit Office (NIAO), as well as other colleagues who are present. Welcome back to the PAC, Neil. I know that it is your favourite place to be. Thank you very much for agreeing to attend, along with Paul, to take questions from the Committee for this important evidence session.

This session follows the publication in July 2025 of the Comptroller and Auditor General's report, 'Major IT projects in Northern Ireland'. That report examined the portfolio of 29 major IT projects, with an estimated whole-life cost of approximately £5·2 billion. It identified serious and recurring weaknesses in how major IT investments are planned, governed and delivered across the public sector in Northern Ireland. The report highlights persistent problems, including significant delivery delays, cost escalations and prolonged reliance on legacy systems that are operating well beyond their intended life. In many instances, existing contracts have been extended repeatedly over a number of years, often at substantial cost to the taxpayer and with increased operational-resilience and cybersecurity risks.

Of particular concern to the Public Accounts Committee is the report's finding that many of those issues arose at the earliest stages of those projects. There were weaknesses in defining the scope and intended benefits, as well as unrealistic assumptions about timescales and affordability. Insufficient capacity and capability to deliver complex digital change have, in too many cases, undermined projects from the beginning.

Today, the Committee will be seeking assurances that the Department of Finance is addressing those issues. We wish to understand what has changed since the report was published by the Audit Office, what risks remain and how Departments are ensuring that future major IT investments are delivered to time, to cost and with clear and measurable benefits for the public and service users. The report also points to wider system-level issues, including the absence of a coherent Civil Service-wide approach to planning for and managing major IT change and developing the specialist skills that are required to deliver that change. The Committee will therefore be considering how those challenges are being tackled collectively across Departments.

As always, Neil, we expect clear, direct and candid evidence from you today. I invite you to make some brief opening remarks. Members have a number of questions in relation to this important evidence session. I ask that you be brief in your remarks and precise and to the point in your answers. We have a lot to get through, and members' questions will tease out the areas of interest for the session. With that said, you are very welcome, Neil, and the floor is yours.

Mr Neil Gibson (Department of Finance): Thank you, Chair. I thank you and members of the Committee for the opportunity to be here to address the report on major IT projects. Robust independent challenge is an essential part of how we improve our services and manage public money, particularly in areas as complex, high risk and business-critical as IT-enabled change. The Audit Office report provides a clear and helpful diagnosis of long-standing systemic challenges in how major digital change is delivered. It highlights several recurring themes and provides useful insights into and observations of the challenges in delivering IT projects, many of which are not unique to this jurisdiction but, nonetheless, require careful consideration and focused attention.

Since the period covered by the report, a number of practical steps have been taken in response to the recommendations, including tighter controls in contract management; cost information; strengthened gateway and independent assurance arrangements; clearer standards for business case development; and an increased emphasis on professional capability in project delivery, digital and commercial disciplines. However, process improvements alone are not sufficient; a more fundamental shift is required from seeing IT projects as technical procurements to recognising them as business transformations with digital, organisational and cultural implications.

The Committee will appreciate that the responsibility for the delivery of IT projects rests with the responsible Department or public body, so I do not have the detail underpinning other case studies that are referenced in the report, or the work that is being progressed in other delivery sectors, such as health or education. However, I am the permanent secretary for the Department of Finance, and two of the five case studies reside in our area. I am joined by Paul Duffy, deputy secretary and senior responsible officer (SRO) for the Integr8 project, which is one of the case studies in the report.

Many of the recommendations in the report apply equally across all Departments and the wider public sector, and I am happy to pick up on areas in which our shared services are relevant to system-wide delivery, such as project governance, project and contract management, skills, business case guidance and others.

The Chairperson (Mr McCrossan): Sorry, Neil. Paul, your mic is tapping the laptop and is affecting the recording.

Mr Gildernew: We can translate the Morse code. We know what you are saying.

The Chairperson (Mr McCrossan): Sorry, Neil. You can go ahead.

Mr Gibson: It is normally your knees that knock with nerves, but that was our laptop.

You will be glad to know that I am reaching the end, anyway. I have kept my opening remarks very brief. In closing, I understand that DOF has an important role in supporting Departments in improving project delivery, whether that is through the shared services that we provide, the outworking of policies or guidance that we provide, or in helping to create an environment in which shared learning is encouraged and facilitated. We look forward to answering the Committee's questions.

The Chairperson (Mr McCrossan): Thank you very much, Neil, for being precise in your opening remarks.

The Audit Office states that the lessons are known and have been widely reported, but projects are often still not set up for success. The Committee has seen the same weaknesses and failures across a wide range of reports: poor strategic planning; capacity and capability difficulties; governance issues; and costly contract extensions. Those issues are not isolated to just one Department; they are systemic. Why have senior leaders across the Northern Ireland Civil Service not yet addressed those long-standing issues?

Mr Gibson: Thank you. Some addressing of that is under way. We have been in front of the PAC before, talking about major capital projects. There are some similarities in how we look at modernising our professional structure, which is under way. That is not something that you can do overnight. There is better reporting, and better data is being provided centrally, but, when thinking about this particular report, one of the key points that we have to recognise is just how complex IT projects are and how different they are from some of the other major capital works that we talk about elsewhere.

In some of the areas that you highlighted, we need to recognise — this has certainly been my experience — the greater challenge in being able to be very precise about what your benefits and cost profile might be when trying to predict technologies that maybe do not even exist today. I will give a very real example: I, Paul or any expert would not be able to tell you what IT might be able to do to help you to deliver, but we have to craft business cases and approaches without that knowledge. That leads you to look at a more modular approach to how to deliver IT projects and at longer-term strategic partnerships. The Northern Ireland Civil Service (NICS) board has taken a leadership role around major capital projects through its investment strategy for Northern Ireland (ISNI) committee, which we talked about previously. IT projects do not sit within the scope of that strategy, as you are aware, but there has been the establishment of the Office of AI and Digital within TEO. Following the digital maturity assessment that was carried out, there are plans to look towards the recruitment of a chief digital officer (CDO), which will probably be set within TEO structure, to look across systems' lessons. That is not a resource that resides within DOF. It is, therefore, an issue that has been raised and debated at the NICS board, although it has been predominantly talked about in the context of the major capital works that we have talked about before.

By their nature, the way in which IT projects — I am sure that we will talk about some of them — interface with other systems and projects can be wildly different between Departments. The lack of commonality across them is striking. We have put in place initiatives — we have corresponded on those — around training for people who, like Paul, have gone through the Major Projects Leadership Academy (MPLA) training; made sure that we have more staff; and reported, through the SRO programme, to make sure that there are greater skills. We are also developing a project delivery profession. There is, therefore, a lot of work under way, demonstrating that the issue has been looked at and is constantly being addressed. However, it is important to state that trying to change a structure or develop a new profession takes time.

The Chairperson (Mr McCrossan): Thank you very much for that, Neil. How would you, in a few words, describe how the Civil Service has managed the IT systems to date?

Mr Gibson: I can only speak with any authority on the systems that reside in my area. I do not have any insight or in-depth knowledge of the projects that exist in other Departments. Delivering its digital projects has been an incredibly challenging journey for DOF. We will come to Integr8 in due course and to some of the decisions that I have had to take during that period that have led to some of the delays that you speak about. Within the individual projects, it has been incredibly challenging in areas, which, I am sure, you will talk about, such as skilling and securing funding, but I am confident that the Department of Finance has a good and robust governance structure to manage its digital budget.

The Chairperson (Mr McCrossan): Someone looking on who has considered the report, its findings and the estimated cost of over £5 billion, will, rightly, be shocked at the scale of failure when it comes to major IT systems. Given Northern Ireland's size — it is six counties, 18 constituencies and 11 councils — why has the Civil Service in Northern Ireland got it so badly wrong?

Mr Gibson: A lot of the evidence in the report on the IT project delivery failure is drawn from figures around cost extensions, for example. Of course, while those projects have been extended, the service is still being delivered. Very often, replacement systems will be more expensive. Therefore, the idea that there would be a financial saving, for example, if those projects had not been delayed, is not quite a true reflection of what would have been the case. Take, for instance, an accounting officer who is looking at the digital projects that they might have. They have to secure funding for them — that is a bidding process, and you have to make sure that you get that approved by your Minister and agreed in the Executive — and then make sure that they can recruit the staff that they need. One of the key challenges in digital projects — more so than in any other type — is that some of the skills that you might need are generic, such as contract management and delivery, but lots of the technical skills will not reside within the service, and they will change over the life cycle of the project. Therefore, I do not think that it is fair to characterise it as us getting it "so badly wrong". Think about the areas that I speak to: Integr8 is replacing the finance and HR system. That is one of the things that it is doing. The finance and HR systems still work and provide their services, and they cost money. It is not that they are not delivering and the staff are not getting the services from that. It is just that they are not as good as they could be. It is important to set that figure in context — it still pays for existing services, but they may not be the modern and most efficient and effective services that you would like to have. If we were sitting here with no delays, the amount of expenditure on projects may or may not be any less than what you are seeing today.

The Chairperson (Mr McCrossan): That is hypothetical.

Is it the case, Neil, that the Civil Service has allowed failing IT systems, which are neither good value for money nor delivering what they are supposed to deliver, to continue because it is easier to do so than enter into all those renegotiations and plan out a system that benefits service users much more cost-effectively?

Mr Gibson: I cannot speak for others, but that is not the case in DOF's considerations. Given the limitations on financing and skills, you have to be sure that you can sequence all your projects at the same time. Integr8 progress meetings are lengthy, detailed and technical and, as an accounting officer, you are going to struggle to have five or six of those running concurrently. You might, in principle, look at replacing all the systems, but that would not make any sense, because you could not do it in the time that you have available.

There are times when new knowledge becomes available, so you might look at replacing something with a new system. You might decide that there is some cloud potential, and AI is the obvious one now. You might think that you are not in the best position to launch a new procurement process, because you know that the technology is moving so rapidly that, in another few years, you might have to assess that calculation differently. You have to approach each one differently and separately. It is not a case of leaving all your systems as they are, but you revisit your decision-making throughout — I certainly have.

The Chairperson (Mr McCrossan): Is it the case, Neil, that by the time we bring some of those systems up to date and to the required standard today, they will be out of date by tomorrow because of advancements in AI and other technology? What potential issues does that present for service users?

Mr Gibson: That is hugely important, and it is why we have to change how we think about systems and, increasingly, look at service arrangements that can be updated continually, much as you would be used to with the technology in your phone or your laptop. When we speak about Integr8 later, we will talk about how the idea there is to have a system that will have patches and updates to keep the technology moving and embrace AI when it becomes powerful enough to help us but will not limit us to one type of system, technology or deliverable. We need to move away from the idea of a "big bang" procurement to a programme of continual system delivery and improvement.

That is another way in which things such as IT projects need to be thought about differently. If you think about that, that tends to lead to longer life cycles for projects. If you went to a contractor for a seven-year contract, but it was going to take two or three years to procure a new one, they will only be in the door when you start to talk about their successor. It is better to think about it more as a system that evolves, and that is in keeping with how other jurisdictions are approaching that issue, looking at service delivery programmes rather than thinking about technology solutions. You are quite right: they are becoming dated very quickly.

The Chairperson (Mr McCrossan): OK. I have one more question for now. You have said that there is a clear need for a strategic approach to delivering IT systems and the necessary digital infrastructure. That is essential to unlock the potential of new technologies, which we have referenced in some of the discussion so far, and to delivering on the digital transformation benefits of the Programme for Government (PFG). However, we have a lack of overarching strategies. It has been well rehearsed in evidence to the Committee that there are silos across Departments and data is of poor quality. Again, that is a theme that continues to present itself in these evidence sessions. As we have mentioned, there is a heavy reliance on legacy IT systems. What role does your Department take in driving a strategic approach to enabling digitalisation and in ensuring joined-up services across the Civil Service?

Mr Gibson: We have a responsibility in that we provide a set of shared services. We look to a content manager for the delivery of those shared services, of which Integr8 is one. There are a number of IT systems on your work laptops. We have a responsibility to coordinate and deliver that across the service. We do not have a central responsibility, however, for the technology solutions that exist in other Departments, their arm's-length bodies (ALBs) or their line of business. Although it might be quite appealing to say that it would be better to have central control of that, previous experience and experience in other locations show that there is a limit to that, because those technology solutions are so widely different across Departments. DAERA, for example, has a lot of interfaces with European systems and has to work through Windsor framework arrangements that would not be relevant elsewhere. The Department of Justice has particular security controls and clearances that would not be needed by other Departments. It is quite difficult to think that there is enough commonality.

However, we have moved, through the digital maturity assessment that was carried out, to secure funding to appoint a chief digital officer, who will start to look at those structures, systems and processes. It is certainly not the case that you could easily say that there is enough similarity to have some central service delivering all those IT projects. We have, I think, 89 legacy IT systems in the DOF family. Multiply that across Departments, and you can imagine how unwieldy it would be to have some sort of central assessment of all that. There is currently a lot of Department-led control, and that is probably still right when we look to the future. DOF's responsibility is limited to ensuring its shared services.

Mr Paul Duffy (Department of Finance): I will add to that. One of the things that DOF can do for other Departments is look at where there are common tools that apply across multiple systems. An example is the identity assurance solution. That is provided by DOF, and all Departments and arm's-length bodies can utilise it. Where there are opportunities for DOF to provide a common solution, we will do that so that individual bodies do not need to have their own. There is an identity assurance system, and there is a payment system. Instead of those systems being replicated across the public sector, there is one for all. The identity assurance system was hugely beneficial during COVID. The Department of Health was able to lean in to and utilise that.

The Chairperson (Mr McCrossan): OK. Thank you for that, Paul.

Mr T Buchanan: The Civil Service ICT strategy 2017-2021 provided direction for investment right across government and a more collaborative approach to commissioning, delivering and managing new technology investment. Why was the strategy not fully implemented? What went wrong?

Mr Gibson: That predates me, but there are a few things. The strategy is nearly a decade old. If you were to look back and read it, you would not think of the IT and digital landscape in quite the same way. Of course, the big event that happened was COVID, which moved a significant re-profiling of digital resources, such as payment systems, to what could and should be done to get a workforce to be able to work from home. That was a distraction.

After that, we had periods of political instability. From a Department of Finance point of view, once we had sufficient bandwidth away from those crises, it did not feel to me, as accounting officer, that an update or a review of the strategy was the most appropriate way forward. It is not the strategy that I would write now. That has led us down the route of looking towards an office of AI and digital maturity. In a sense, the world has moved on a little from where that strategy would have been. Reading it back, I think that it would have been impossible to deliver some of the directions that it sets out without funding a service delivery agreement (SDA) and greater funding, which simply was not available. DOF does not have that funding in-house. There was not the ability, therefore, to put that kind of resource towards it. However, I do not want to play it out as, "It hasn't been resourced". Looking back, as you would expect, 10 years hence, I do not think that you would start from there.

In a sense, we have parked that and moved towards looking at a CDO, a different structure, a sprint mentality and a different way of thinking about it. Interestingly, there has been very little clamour from any Department to me, in the time that I have been in this role, to seek to revisit the strategy. That has never been demanded. People have not said, "If only I'd had that strategy, I could've done something better.". Some of the principles that it mentions, such as the approach to digital and a shared vision, still hold, but it has not been possible to implement that in the way that it was envisaged then, because the world has changed, and the financial situation has not been favourable, to put it that way.

Mr T Buchanan: With no current Civil Service IT strategy in place, is the Civil Service delivering as it should? Is it delivering as effectively as it should?

Mr Gibson: That is a very difficult question to answer in any precise way. I do not think that the absence of a strategy is holding us back in progressing those line-of-business systems that we are doing. It certainly has not held us back in the development of Integr8 or any of the ones that I am responsible for. I would not sit here and say that I feel the lack of a strategy has materially held back any level of delivery and development. I do not think, though, that I would ever sit in front of this or any other Committee and say that we are doing as best we can. You can look around and see that, in almost everything, you would wish for more resource, more people and an ability to move things faster than you do currently. We take some comfort from looking at other regions and talking to colleagues in other jurisdictions about just how challenging the digital transformation journey is. Where we can, we are trying to pick up on their expertise. We have UK members sitting on our Integr8 board. We are trying to learn from their experience. They are certainly not going towards wholesale ICT strategies across the system; it is too complex and too diverse. We are looking more at a set of consistent principles and proper project management in each Department.

Mr Duffy: Some of the larger Departments, such as DAERA and DOJ, have individual digital strategies, which are very much tailored to their particular services and needs.

Mr T Buchanan: Why is there no overarching strategy in place?

Mr Gibson: There are a lot of reasons. The resourcing of it is one. It is easy to talk about the high-level principles that you might have. A chief digital officer, when in post, would certainly look at the merits of that. You also have to have the permissions: who can make decisions on behalf of other Departments and the IT solutions that they may wish to progress with? You need to have a common understanding of the powers and remit of any new role, and you need the funding that goes alongside that. However, as I said, each digital project is really different, so I am not confident that a centralised strategy would be a key missing piece in the direction of improvement that we are all looking for.

The Chairperson (Mr McCrossan): Before we go any further, you mentioned consistent principles. What are the consistent principles in the absence of the strategy?

Mr Gibson: A Department that has its own strategy might say that there are certain things that are in principle. We can think of the very straightforward ones that every Department will have in every project, such as security of data, and interoperability, where you can, with other systems, but there will be nuance in the key priorities for each Department; they will differ. I can speak to only ours. We set out a set of principles for our strategic projects.

The Chairperson (Mr McCrossan): The Audit Office has done a brilliant job of pointing out one very real consistency, which is that the entire Civil Service has handled the whole thing terribly. It has been a disaster.

You mentioned resourcing. Neil, I do not think that people will accept what you said about it coming down to resourcing. The cost of updating the current infrastructure will be in excess of £5 billion, and it is rising by the day. How is it acceptable to say, "We'll put a sticking plaster on this", instead of funding the transformation to bring the system up to date, so that it meets the needs of the people who need it?

Mr Gibson: I do not disagree with that statement, but there is a budget process. You have to bid for the money that you need. You have to identify the skills that you need. You have to set up permissions; you might have a central unit or team that has responsibilities in other Departments —

The Chairperson (Mr McCrossan): How can you do that without a strategy? If there is no strategy that points to that solution, how are we going to achieve that?

Mr Gibson: Looking back at the previous one, do I think that a written strategy would do anything material to improve the delivery of the projects that I oversee? The answer would be no, so I wonder whether producing another strategy document would be the best use of resources.

It is important to state again that there are savings to be found. There are new systems to modernise and change, but the history of all jurisdictions is littered with IT challenges in delivering new projects, and new systems being more expensive than the ones that they replaced. We will come onto this much more significantly later, but, for almost everybody who interfaces with a system, be that the stuff that we provide to our staff, or the stuff that citizens engage with, the programme of change for them is enormous when it comes to communications, training and upskilling. The amount of resistance that you get to any form of change, such as a new system or a new way of working —

The Chairperson (Mr McCrossan): In the Civil Service?

Mr Gibson: Anywhere; in any system in any —

The Chairperson (Mr McCrossan): That is part of their job. With respect, Neil, I do not think that the public will accept that. In fact, they will be infuriated by it. There are a huge number of people working in the Civil Service, paid by the taxpayer, who are required to deliver the necessary changes. It is not acceptable that they are uncomfortable with it; they are supposed to move with the times. Is there any desire at all among people in very senior positions in the Civil Service — one of which you hold, Neil — to move to a joined-up digital service for the public on the ground, at the grassroots and at the coalface, who do not care what Department delivers the service? A strategy is needed to join up those services.

Mr Gibson: I cannot speak for the other senior leaders, but I think that we all aspire to the idea of a digital-first, simpler interface with public services.

When you sit where I do, you are much more aware of the complexities involved in moving towards that and of the risks that you take when you say, "Well, this would be so much better if it were just done digitally or if it were in the cloud". The efficacy of data has to be protected. It is a more careful journey than it might sound to the public and is not just a case of saying, "Put everything in one system and on the one route". There are challenges to be faced around data-sharing agreements and legislative programmes. I do not disagree with the sentiment, but, sitting in my seat, I have a clear understanding of the reality of how challenging it would be to deliver that.

The Chairperson (Mr McCrossan): Sorry for interjecting, Tom. A few burning issues came up there.

Mr T Buchanan: It is a wee bit disappointing to hear you be so dismissive of having a strategy. Without a strategy, all the other Departments will continue to work in their silos. We have seen that that does not work, and it has not delivered. What is being done to ensure that the IT systems and platforms that are procured today are designed to accommodate ever-changing and evolving technologies?

Mr Gibson: I am sorry to have disappointed, but I do not see the level of commonality that is perhaps envisaged here, which is the idea that one central Department — TEO, DOF or wherever, you think, it would sit — would have the level of detailed expertise to understand what type of system DAERA, for example, might need for animal movements or farm payments, when it might be a completely different technology platform that has nothing to do with it.

The best that I can offer is the way that DOF thinks about such things. We think about how we provide services to other Departments. We do not have any remit, powers or control, and I do not know that it would be entirely appropriate for us to have that, because there is a lot of specialist expertise in Justice, Agriculture and Economy. Those Departments know what types of systems work for them and their suppliers, so they have appropriate strategies for their work. I am not sure that a central strategy would set out that they should all work in the same way. Many of the systems with which we have to interface are not in our control — systems that the EU or GB use — and we could not decide, unilaterally, to go in a different direction. It seems appropriate that each Department assess that.

Now, could we look at each Department's digital strategy and think about whether there are common themes and approaches? Yes, but I do not think that it would be a detailed strategy where everyone should follow the same type of system and delivery. That would not reflect the complexity in each Department.

Mr T Buchanan: I have one more issue, Chair. How do you propose that Departments break out of their silos and work collaboratively?

Mr Gibson: A few practical things can be done to reduce silo working, the most obvious of which is to have greater networking and communication. We have a senior information risk owner (SIRO) forum, within which there is a cybersecurity group. We make sure that we have an SRO for all the people who lead the major projects and bring them together as a community. There is a Teams channel through which they can engage. We run events to try to do consistent training. We try to create a family and a network so that people will feel that somebody else in the service will have gone through something similar. We draw in expertise to sit on our boards, panels and delivery programme boards from people who have been through it before in other jurisdictions. The biggest breaking down of silo working comes from building up relationships and engagement. We have set up a number of platforms to do that.

This is a very minor point, but it is important. As we move through the delivery of Integr8, we will get better and better data about our staff and set out better career pathways for them when they are involved in project delivery or in the IT or digital professions. That will make us better, as an employer, at attracting staff and help us to do the job that we will have to do, which is to train our own —

The Chairperson (Mr McCrossan): When —?

Mr Gibson: — to meet those needs.

The Chairperson (Mr McCrossan): Sorry for interjecting. When will that happen?

Mr Gibson: The job families work is going on at the minute. Integr8 will come in, in phases, from next year. You would not design a new system to go into the old one; you put it into the new one. We will be rolling it out over the next few years and will have all those new systems in place by 2028. The work to get the new professional structure ready for that system is ongoing.

Mr Wilson: We understand that an AI strategy and a road map for AI adoption are in development. Will you give us an update on that work so that we can determine where AI could add the greatest value and deliver efficiencies, both in Civil Service processes and in public-facing services?

Mr Gibson: Thank you. You may wish to communicate directly with TEO for the detail on that. That is where it resides. I am not over the detail, although members of my team engage with that. I do not have any detail on the AI unit, as it does not sit in DOF. However, we have been rolling out some AI-enablement tools and making sure that our shared services can avail themselves of those. Paul will want to speak to some of the tools that we have already rolled out.

Mr Duffy: I will give you some real-life, practical examples. One thing that we have rolled out to all Departments is the ability to use Copilot to assist their staff with their jobs. That is just one AI tool within the Integr8 programme. As part of that solution, we are developing an AI chatbot to allow people to search for and find information on policies, processes or ways of working. We are facilitating a number of AI tools that are more common across Departments, but, in the future, you will start to see that AI will be built into existing applications rather than being stand-alone tools. As Departments start to procure and develop new systems, the AI capability will be built into those systems and will be available to them.

Mr Gibson: We use some AI tools in the Northern Ireland Statistics and Research Agency (NISRA) to help to rewrite code and modernise that, so that we do not need to do surveys in the old way that we used to do them. We can automate the results that are published. There are lots of examples of where AI is already playing its part, but the AI strategy will also consider those matters of efficacy, cases where data is protected and where the information ultimately comes from. I encourage the Committee to get an update directly from TEO in case I have missed anything. The actual unit does not sit in DOF, but, as you can see, where we can, we are rolling out shared AI services that people can use in each Department.

The Chairperson (Mr McCrossan): By the time that we get a response from TEO, the system will be out of date. [Laughter.]

Mr Wilson: I want to dig deeper into AI from the Civil Service's perspective. What type of protections do you envisage there being for staff where questions are asked and AI is perhaps being relied on too heavily or not enough? How do you see that being benchmarked, so that the type of information that we get has a degree of individuality and is tailored by the expertise for which the person has been employed to deliver, as opposed to someone just typing something in to get information from a bot and ignoring their particular expertise? In the Civil Service, we see bum-covering to an extent. Will we see AI being used as a scapegoat at times, where people say, "Well, AI came up with that", as opposed to an individual?

Mr Gibson: I do not want to speak out of turn, as it is not my area of expertise, but that is probably one of the biggest challenges as we face into AI. For me, it starts very much with being open and transparent about when AI has been used and for what purpose. It can be a really important tool in training and development, but it is collating information that is out there, and that information still has to be produced. It does not take away from the need to create primary information.

Mr Wilson: It can be wrong as well.

Mr Gibson: It can be wrong. There can be hallucinations and all the other things that they talk about in what AI produces, such as false positives and so forth, so care has to be taken with AI. At the moment, however, its key abilities are to automate fairly routine processes, but it is moving very quickly towards being able to do other things.

Take recruitment, for example. It is hard to imagine that any people who are applying for jobs within the next 12 months will not be using AI in their application form, if they are not doing so already. That creates challenges for how you assess that. You have to give guidelines — we have done that — in policies on what you cannot use AI for.

I would not say that it is my area of expertise, but you have hit on some really important points about what data you make available. At the moment, you cannot interrogate a number of our systems with AI tools, because they are held on bespoke systems. Data becomes a very important part of that.

Mr Wilson: From what I have gathered from you so far, you are not particularly struck by the adoption of cross-departmental strategies, but it sounds like an AI strategy would be useful in that particular instance. It is a concern, and I understand that you cannot know everything in your current position, but there is a lack of a strategy. How will we achieve a strategy that encompasses all Departments, sets some benchmarks, protects the essence of the services provided by the Civil Service and delivers for ratepayers and people who rely on government services?

Mr Gibson: Lest I upset lots of my staff, I had better not be too anti-strategy. It is about whether it is appropriate to tackle a particular problem. With something such as AI, you may want to have common principles, approaches and standards that you might want to apply. The Office of AI and Digital is looking at that. I have less certainty on the issues that, I think, are so disparate that an overarching piece would not bring it all together in the same way.

Mr Duffy: I want to give you a little bit of assurance around that. There is very clear guidance for staff on how to use the Copilot AI tool, which has been rolled out across the Civil Service, and there is human oversight. What it produces has to be checked and cleared by an official, and there is training and support for staff in using that. All the data that is put into it is held in a secure environment. You cannot put something in that will appear on the internet somewhere. The data is locked down and secure.

Mr Wilson: The use of AI can be shaped by opinions, global events and, as you said, what is input to it. Civil servants then rely on that. We have seen instances where perception has been changed by the use of AI, and facts can be ridden over. My concern is that, without a strategy and with a potential reluctance to have a strategy, we may see problems ahead with the use of AI across government. That needs to be monitored very closely.

Mr Gibson: Yes. I definitely consider the approach to AI to be a different thing from a strategy to deliver similar ICT projects across Departments. There is much more merit in that type of approach. You have raised the question of very important risks. Let us be honest: what it can and will be able to do will be hugely advanced in another 12 or 24 months. We would need not to take 12 or 24 months to produce a strategy.

The Chairperson (Mr McCrossan): I will follow on from a couple of points raised by the answers to Gareth's questions. From the outset of the session, you have said very clearly that you recognise the importance of AI. I would be more worried that you were not at the races on the technology stuff had you not said that. You also said that AI sits with TEO, so you do not have too much detail on it. That gives me some concern, because, as the Chair of the Committee, I consider you, as the accounting officer of the Department of Finance, to be one of the most senior officials in government here. It worries me that you do not know what is happening in relation to that in TEO. Does that not prove our point that there is no joined-up thinking and that there are missed opportunities all over the place, the end result being a huge bill for the taxpayer?

Mr Gibson: No, I would not necessarily accept that.

The Chairperson (Mr McCrossan): Which part?

Mr Gibson: I have an odd feeling that there is a level of knowledge that, you think, I could have in my position of seniority. I will be straight with the Committee: I do not have a big enough brain to retain all the things that the Committee would like me to be able to retain. I rely on other people to have that expertise and knowledge. AI is moving so fast that what I have to take care of, which Paul has spoken to, is that, when we are deploying it, we make sure that we have the right standards around the data and that we ensure the efficacy of that data and its protection and security. When we think about it centrally, we put our faith in the AI and digital unit and what that will do and what it will produce. We will see what the strategy brings and look at how to deploy that. There is, however, a limit to what a senior accounting officer in the Department of Finance can do to pass themselves off as some sort of AI expert.

The Chairperson (Mr McCrossan): I am not asking you to be an AI expert. We all learn daily about how AI has the potential to reshape, for good or bad, our way of life. In numerous answers now, you have said that, currently, the AI team sits within TEO, and you do not have any level of detail. Given the significance of the technology and the criticisms of current IT systems in the Audit Office report, I am worried that you have not had — I may be wrong — conversations with the accounting officer or the team in TEO about what is happening with AI. It will have a direct effect on how your Department operates.

Mr Gibson: Yes. Paul alluded to the conversations that we are having in the Department. We have had the conversations about the roll-out of Copilot and how we use the AI tools that we use. We have thought about that through our HR policies and how it is used or not used in recruitment. In the areas where DOF has a direct relevance, I would say that, as accounting officer, I have ensured that I have visibility of those. What I do not have is some of the blue-sky thinking around the potential of AI that the experts in the AI and digital unit might be thinking about and putting together. Like all accounting officers, I will look at that material when it is produced.

The Chairperson (Mr McCrossan): The worry that I have, Neil — I do not think that I am going to get the answer that we would like on this — is that typically, as with most of the issues that present to the Committee for investigation, there is a complete lack of joined-up thinking, absolute silo working and no interest in what your neighbouring Department is doing. That seems to be the theme when it comes to the IT projects. I am more frustrated at the beginning of this session than I have been for quite a few, and I have been very frustrated with them, that there does not seem to be an appreciation of how important it is that our systems are up to a standard that meets the expectations of the public and the value for money that is expected for that. All I am hearing now, and I may be proved wrong before the end of the session, is that it is business as usual.

Mr Gibson: I do not have an answer to that.

The Chairperson (Mr McCrossan): OK. We will go to David.

Mr Honeyford: I will follow on from that and stay on AI. The Office of AI and Digital was established last year. Talk us through how that collaborates right across the Departments.

Mr Gibson: It is in its infancy, so there is not a lot of material coming out of that, which has come to the NICS board or is being deployed across Departments. It is in its early days. You would need to write to it to see what its work programme is. That will come forward, and we will deal with whatever it brings and says. I do not want to speak out of turn, because I am not close to its day-to-day workings, but it will certainly be looking at the points that others have raised about the processes and ways in which we need to think about AI across the system to be consistent right across the public service. You would be best to go directly to that office.

Mr Honeyford: That office opened in June 2025. In December of the same year, so five or six months later, the Department for the Economy announced its own team and launched its own strategy. Therefore, there are two strategies in two Departments, six months apart. Is that not absolutely a case of silo working in the Civil Service?

Mr Gibson: You would have to look at the detail of both of them to see whether they are talking to the same point or same thing. Each Department is large, and its stakeholder group is enormous, with different groups and different people, so there is not always quite the commonality across issues that you might think that there is. I have not read the DFE one in detail, so I do not know whether there is any overlap or similarity.

Mr Honeyford: My difficulty is that you have said, a couple of times now, Neil, that nobody does it with a central thing. I understand that the Department of Health will need a specific thing, as will DAERA. I get all that. However, ultimately, if you look at any other jurisdiction, you see that they have a central framework with spokes out. The Department of Health, DAERA, Justice, Communities or whichever Department will have its own bespoke one, but the principles, guard rails and ethics around all that will sit centrally. We do not have that.

Mr Gibson: No, not at this point, but the Office of AI and Digital will be looking at that component. The digital maturity assessment suggested that a chief digital officer would look at the digital and IT transformation in the same way, and those would be, in the way that you have described it, a kind of central model. However, as I have said, the AI office is only in its infancy.

Mr Honeyford: The most frustrating thing for me, sitting on this side of the table, is hearing, "We are going to look at or talk about that" or, "This will potentially happen at a point in time". Honestly, every time we sit here, that is what we hear. We are here, looking backwards, to ask why that has not happened. We are looking at billions of pounds of overspend. Why, in 2026, has that not already happened?

Mr Gibson: OK. We have talked about some of the things. We have rolled out lots of things that have brought about savings and developments in our Department and in the shared service that we have responsibility for. I struggle, in a practical sense, at just how much time and bandwidth you could have to lean into the idea that everything needs to be done centrally. It is just not how I have experienced the individual IT systems. We have done Integr8 entirely without silos, completely across all Departments, and that is a project specific to what we deliver and have the competency to do. I am not sure that the corollary is to say that, if you had something central that was delivering, it naturally follows that there would not be some of the challenges or things that we face. I am not sure that that is clear. Hopefully, when we come onto it, we will talk through the entirely collaborative, non-siloed approach to the new enterprise resource planning (ERP) system, Integr8. I hope that that is the sort of model that will be held up as an exemplar. I am not sure that that missed out on having something centrally to control it, because that is a shared service that we deliver. I am not sure that we would benefit from having greater insight into a line of business service in another area. I am not convinced that there would be huge benefits from the extra funding of some sort of unit or group.

Mr Honeyford: The South has gone to a central AI model. Wales, Westminster and Scotland have done exactly the same, but you think that we can do something different.

Mr Gibson: No, that is not what I said. The office of AI is already established, but I am not responsible for it, so I cannot answer questions about it in any detail. Again, a chief digital officer is not within my remit. It is very difficult for me to answer questions about what might be happening with something for which I do not hold responsibility.

Mr Honeyford: You talked about the role of the Civil Service board. Ultimately, what is the role of the head of the Civil Service?

Mr Gibson: We all have roles as senior leaders. I am not abdicating from the fact that thinking about what needs to be done centrally is something that all senior leaders in the Civil Service should consider and contribute to. We, as accounting officers, meet every Friday, for example. We have considered and agreed the structure of the future chief digital officer and what that role might look like. Those discussions happen. The NICS board has three committees on areas that you have looked into before: people, major capital and the PFG. There is a bandwidth limit when it comes to time and the number of issues that it can cover. We sometimes pick it up on a Friday at the permanent secretaries' stocktake (PSS). The structure of the AI and digital unit has been discussed in that group, as has the future of the CDO, so it would be completely wrong of me to give the Committee the impression that there is not central thinking and looking at learning best practice from elsewhere; I am just limited in the detail that I have of that to give you.

Mr Honeyford: You have said, "We're talking about it" and, "We're discussing it". People want stuff to be delivered. A serious amount of taxpayers' money is being wasted. It is not there to be spent on central services. We are told, "The Civil Service is talking about it" or, "We're discussing it. We'll maybe bring it up". We need action and for stuff to be driven forward. Every other region in these islands is making progress, but we are still talking about it.

Mr Gibson: I am sure that we can quote a few figures about how other jurisdictions are finding the delivery of their projects, which do not stand out as any different from ours, so I am not sure that the corollary is that, because we do not have it, those things follow. We listed some things that we have not spent any time on discussing here. With virtually no team, we have rolled out the use of Teams and Copilot. The idea that we are not rolling out new technologies and ways of working or delivering major IT projects, some of which we mentioned, is wrong. Those things are happening, and they are the same things that are happening in other jurisdictions.

Mr Honeyford: You are doing that, but Economy is doing something else, and Health is doing something else again. There is not a central framework, with an ethical bit behind it, which everyone else had to have, from which you allow everything to roll out. Everybody is scattered, doing whatever. There is not a central basis to it, with ethics or values. Everybody is thinking again, so it is being repeated in all those Departments.

Mr Gibson: I am sure that more will come out of both of the offices to that effect. As Paul mentioned, where we provide a shared service to all, we ensure that that does not happen. We make sure that, when it comes to what is within our control, people use the same systems in the same way.

Mr Honeyford: That is the issue.

The Chairperson (Mr McCrossan): Neil, to follow on from a couple of those points, we have heard quite a bit about data-sharing. It has been an issue in numerous inquiries conducted by the Committee. There are also issues with central coordination functions and strategies. Where does responsibility lie for ensuring and managing data-sharing and agreements?

Mr Gibson: Different Departments have their own arrangements. Large ALBs also have their own; there is no central control in that regard. We have responsibility for our own network, the data that we hold on that, and relationships and the data-sharing agreements. Some of those require legislative cover, depending on whether they interface with other public data. Each Department looks through its own lens at the particular data that it holds.

We are responsible for the data that we commonly hold across the service, which we have a number of processes and procedures to cover.

The Chairperson (Mr McCrossan): That is grand. OK.

Mr Burrows: Good afternoon. I have also had some IT issues, so I am working off my phone.

There are issues with legacy systems. In contrast to England, where the legacy IT risk assessment framework has operated for several years, there is a lack of understanding in Northern Ireland of the extent to which legacy IT systems are used across the Civil Service. What is being done to address that?

Mr Gibson: We are working on that. An assessment of each Department's legacy systems is being carried out. Just last week, we were talking about the DOF's assessment of its own systems. Each Department has a register of when its contracts come up for renewal, and assurance work is being done now. My Department has looked at its systems and assessed them on a risk framework, which allows us to put mitigations in place. Paul, I know that your previous team worked on that.

Mr Duffy: Yes. The Department of Finance has completed its work on that. We have looked at all our legacy systems and calculated the risk associated with each one. That work has been discussed at our departmental board, and we have actions to address the highest-risk projects.

Mr Burrows: What is the risk?

Mr Duffy: It depends on the system and the solution. I will give you an example. Two of the systems rated as high-risk are associated with Account NI, which is our system for payment to suppliers. That is probably our highest-risk system. The solution to that is the Integr8 programme.

Mr Burrows: If the risk materialises, what will it look like? Why is that a high risk?

Mr Duffy: For example, in 2023, we had to take the system offline to upgrade it, which is very unusual for such a critical system. The fact that it cannot be updated online in a modern way illustrates how old the system is. Taking the system offline for two weeks meant that payments did not go to suppliers, including small businesses, in that time. It is critical to support the economy.

Mr Burrows: That was a planned event —

Mr Duffy: Yes.

Mr Burrows: — so it would be exponentially worse if it hit you out of the blue.

Mr Duffy: We have mitigations in place, including backups and disaster recovery procedures, and we have suppliers on call should an issue arise. We have tested the system's going down and what we would do to restore it. Lots of work has been done to mitigate that risk.

Mr Burrows: OK. I just have a couple of follow-ups. In the absence of a clear framework to identify legacy IT systems and assess the associated risks, how are investment decisions made to ensure the best use of resources?

Mr Gibson: Now that the legacy system assessment is in place, each Department will look through that, look at its level of risk and assess the cost of replacing its systems. You start by asking, in principle, "Do I need this system?", and where you can, you look at ways of modernising them and bringing them into systems that have better controls or are more up to date. With that volume of legacy systems, the first obvious question is whether you are sure that you need them. Your first question should be, "Do I continue with this or can I merge it into something else?" Once you have assessed the risk, you have to look at mitigation strategies. Paul talked about some of those. All Departments have agreed with that recommendation and will go through all that. There may be some common themes in the types of system that have come up as high-risk.

We also have to assess the expense that can be associated with very particular legacy systems. In the Department, we have financial databases that we run to look at the Budget. They can be very bespoke and useful only to you, so there are not a lot of synergies. What we are increasingly trying to do — this should be a common theme across Departments — is, where we can, adapt our way of working to suit a more off-the-shelf system so that we can get the support and the security and other updates that are needed, rather than have bespoke systems. That creates a different challenge for Departments, because it is about behavioural or cultural change. We always go into things saying, "I want it to do what it already did, but faster", when that might not be the right way or the way that other places approach it. It is not just a case of saying, "That system is out of date, let's replace it". We should be asking a more philosophical set of questions at the beginning. Why do we have a bespoke system? What is the cost of it? Could it be merged into other systems? Are there technology changes coming that would make it redundant? The first step is to have that assessment. DOF already has that, and all Departments will have that over the coming months. Some already have and are taking action.

Mr Burrows: OK. Two more. Most of the major programmes and projects in the IT portfolio are to replace legacy systems. The Audit Office reported that, on average, contracts were extended by almost eight years. Those continued contract extensions came at a significant cost. The Committee has previously highlighted the risks associated with ongoing contract extensions and the need for strong contract management and strategic planning. Your Department has several substantial ongoing programmes and projects to replace legacy systems that expired between four and 13 years ago. Why are you not learning lessons?

Mr Gibson: A couple of reasons. Replacing a system is complex, as is running a procurement for it. You need to make sure that you have the personnel in place, you have to be sure that your direction of travel is the right one, and you have to have the funds in place. Integr8 is a real example of that. When I came into this post, the Executive collapsed and I had no surety of funding for Integr8 at all. There was no budget agreed for me, much as we are operating today. That meant that I could not commit to Paul and the delivery of that team, and we had to pause that. I took full responsibility for that pause, but the outworking of that was that we had to extend some contracts with existing suppliers. That is just one example of the reasons that arise. COVID was another. The war in Iran might change the cost profile of a project, and that might make you say that you are not ready to go live with it.

It is also worth pointing out that, once you go through discovery, you can find out that the problem is more complex than you thought. That is very common in IT projects. When you start to peel back the layers, you think, "Well, I would not replace it with the same system". You start to look at it, and if you want that extra functionality, that turns into a very different scale and size of project. I have seen that in our Land and Property Service (LPS) projects — for example, the rates and benefits one — where what we thought we would do at the start started to evolve, the solution that we thought would be the best one did not really work, and we moved to a much more modular approach.

There are lots of lessons that we can learn and are learning. It is important for me to articulate one of the big ones. In some of my early challenges around IT delivery, I spent a lot of time arguing, disagreeing and negotiating with suppliers about very specific items in contracts — "Is this in scope? Is it not? Is this a contract variation?" There were lots of lawyers and lots of money spent on negotiation. That taught me that maybe those types of very precise, line-by-line contracts do not work as well in the IT space, where you need more of a partnership. You will say, "You will have a number of development days. I am not sure what they will be used on". That has been a big lesson for me, which we have been learning. Systems change can be complex.

The final one that is important. As we look at IT delivery, the idea of the length of the project —. To me, they almost always look like they would be better longer. I know that we normally do not like to think about that, because that makes it —. Is it still competitive? Have we given the market a chance to compete? In order to get the real benefit from a system, it takes a long time for a project to gestate, to build in and exploit its capabilities, and to build up a relationship with your supplier. So what lessons have I learnt? We need more frameworks and less precision on the exact contract, although that sounds counter-intuitive to what you would normally want. Recognise that longer deals or contracts may be more appropriate in this space, because you are looking for a service and a partnership rather than looking to buy a product.

Mr Duffy: That is a really important point. Were you to award a contract today for 10 years for a system that had just gone live, given that we have said that it takes six and a half years to deliver a project, you are already having to start to think about delivering a new one by year 3, so you are not really getting the benefit out of that partnership. The Audit Office report quoted that period of six and a half years. The UK Government (UKG) take 8·4 years to deliver an IT project. The projects are complex and challenging and are not something that can be delivered in two or three years.

We are now moving away from large system implementations. The issue of having to replace systems every 10 or 15 years is disappearing very quickly. For example, the Integr8 solution will be, in a sense, a piece of software that we are buying, not a system with a licence attached to it. As that system goes live, the software will be updated every quarter with new functionality. We will benefit from that, because we will be in a period of continuous improvement, getting new functionality. We talked earlier about AI. The AI in the current solution is pretty native. In years three, four and five, that AI capability increases, and we will benefit from that, because we will get new functionality every quarter. You are going to get solutions will last longer, but without huge system implementations.

Mr Gibson: The question would then be about how often you would want to go out and contract to change the underlying technology that delivers that platform. Moving from one supplier of an underlying technology platform to another would be such a significant change that doing it within a six- or seven-year window would not really make any sense. Those are some of the lessons that I have taken from looking after projects.

The Chairperson (Mr McCrossan): Which of the current IT systems have overrun by the longest and for what period?

Mr Gibson: I do not want to misquote anything. I can come back to the Committee on that. We have some long-running ones.

The Chairperson (Mr McCrossan): You will have an idea of it. Is it 10 years?

Mr Gibson: Yes, a couple of systems in LPS have been running significantly. I am not sure, but one of our HR systems is quite —. I will come back to you on that, so that I do not get it wrong. My argument is that —.

The Chairperson (Mr McCrossan): This is important for context, Neil, because we are on the public record. I know of people from across various sectors, including schools and other things —. I could speak about Classroom 2000 until I am purple in the face. Colm, you said what, 18 years?

Mr Gildernew: I think that it is 18 years.

Mr Gibson: That is the LPS one, I think.

Mr Duffy: It was DAERA.

The Chairperson (Mr McCrossan): Eighteen years. Dubai was built in less time. Let us put these things in a bit of context. It is just appalling and beyond belief and deeply frustrating. Can we be mindful that the public are watching? I am frustrated by sitting here and listening to some of the answers that I am getting from senior officials. To be very clear: the current situation —. To reset the mood in the room, the current situation is unacceptable. No questions. Do not justify it at all. We are asking why there is no plan to fix it, and we want answers for the public. There is a bill in front of us for £5·2 billion or more. That is the context. The system is a disaster.

Mr Burrows: Learning lessons can save us millions. What can we do to make sure that the lessons that you have learned and what you have done become part of the corporate memory of the Civil Service? Were you to be run over by a bus — hopefully not — or were you to decide to get a better job tomorrow and say, "Here, I am off", how do we make sure that those lessons are learned by someone else?

Mr Gibson: I am not sure that those things sound so bad right now. It is important that lessons are learned. We have the SIRO forum, where we share experience. If we think that there are significant key lessons to be learned, those go through to the NICS board. We have a network and a community of IT people. They meet regularly and share those recommendations and experiences, alongside the informal sharing at engagements and at permanent secretary meetings etc. We have formal mechanisms for sharing and passing through any lessons learned and recommendations.

Mr Duffy: Lessons learned from projects are captured and published by DOF on an intranet site that all Departments have access to. That information is shared, but we also look beyond what is happening in our own Departments. For the Integr8 programme, there are two critical friends from the UK Government who have been through similar journeys to the one that that programme is on. They are sharing lessons learned from their experience. The Integr8 programme team has engaged widely with other jurisdictions, including the Scottish Government and colleagues in the South, to learn lessons. We are not just looking within and sharing but taking from other areas.

Mr Burrows: This is my last question, and then I have an observation on culture that I want to ask as a question. A continued reliance on legacy systems results in lost opportunities to realise benefits that could be exploited by a more modern technology. Where is the innovation and transformation that we repeatedly hear being talked about in the Civil Service? Where is that being delivered when it comes to IT systems?

Mr Gibson: There are a few points to that. That type of transformation often comes from how you exploit a new system, the training that you provide around it and what you do with the new data and capabilities. It is like almost anything that you have on your phone or your laptop: you have to be able to exploit that. It is about the training that you roll out, the use cases and the case studies that you provide. That is where you really see the transformation. It is not necessarily just in the technology. We have rolled out an ability for people to pay their rates bill online. That is an important and much-needed technological development, but the key part of that is the communication of it — making sure that people are aware of it and that the system can handle it.

When the transformation board, for example, looked at its first batch of projects, it saw that so many of them were digital or had a digital component, so it asked for a digital maturity assessment, which looked at how our systems were approaching that and whether there was a lack of commonality. It made some very important recommendations, including on the appointment of a chief digital officer, which is now under way.

Mr Burrows: OK. This one is on culture, and it might be an uncomfortable question. I wonder whether a lot of the problems in the Civil Service around delivery are linked to the lack of accountability. I will give you one fact and one stat. This is not a criticism of the individual; it is about the officeholder. I asked the head of the Northern Ireland Civil Service, who earns more than the Prime Minister — there or thereabouts — a simple question: to whom is the head of the Northern Ireland Civil Service accountable in terms of their performance? The answer is nobody. Zilch. Nada. Nobody. Their wage is substantial; it is about £250,000 or whatever, I do not know.

The Chairperson (Mr McCrossan): Jon, he cannot answer that.

Mr Burrows: I know that he cannot answer that. It is an observation. It is not about the individual. I am going to say the second one. When I asked the question about accountability in the Civil Service and how many of the 24,000 hard-working staff — most of them are very hard-working — had had any kind of performance intervention and been told that there was an issue around performance — I was told that it was 12 per year. The stat is something like 0·00%. I have been in organisations a quarter of the size of the Civil Service where 20 people have been sacked and maybe 10% have been told that their performance was a problem. I wonder whether a lot of this is to do with the lack of a culture of accountability.

Mr Gibson: I do not have a direct answer to that, except to say that I have an accountability framework, including to here and to my line manager, who is the head of the Civil Service. It will be for others to assess whether I have done the job that I was asked to do.

Mr Duffy: I will tie that back to major IT projects. I am an SRO for the Integr8 programme. I have an SRO appointment letter that Neil issued to me that lists my responsibilities that I am accountable for the delivery of.

Mr Burrows: I get the "answerable to", but I am not sure that there is a culture of accountability. It is not about any individual. I am trying to get at whether a culture exists, because I wonder why things are not delivered.

The Chairperson (Mr McCrossan): There is no question that there are frustrations around how the Civil Service operates. However, we have to focus on this particular inquiry today. Thank you for those questions, Jon; do you have any more?

Mr Burrows: No.

The Chairperson (Mr McCrossan): OK.

Mr Gildernew: I am going to move on to another issue around capacity and capability — issues that continue to have an impact on the design and delivery of all major projects, including IT. During our inquiry into major capital projects, Neil, you told us that it is challenging to secure the right skills to deliver capital projects effectively and efficiently. However, on that occasion in April 2025, I think, you advised that you were confident that you had put a framework in place to give you the best chance of doing that. Why are we seeing continued failure? Do you have an update on that framework, and what difference has it made?

Mr Gibson: Thank you, Colm. I contend that we are making good progress. The problem remains partly because the nature of the problem is that it is very challenging to be able to acquire those skills, not just for the service but for the suppliers that we use, which will struggle to necessarily have some of the deep technology skills that are required. We are making great progress on what we spoke about before. We can provide you with the numbers of people who have gone through the relevant training programmes on delivery. We are moving towards a new delivery model and creating all the job families within that. That is incredibly important to allow staff in the Civil Service to look at their prospects of joining, for example, a delivery profession. They can see a road map, and we can look at what type of training we require. The SRO events have been very well attended. So I am very confident that we continue to be on that pathway towards what I spoke of then, but the truth of the matter is that there is a finite supply of those skills. Our nearest neighbours in the Republic and, indeed, our UK colleagues, who pay much more in much bigger economies, cannot secure some of the critical skills within their economies. That means that we have to have tremendously good relationships with suppliers, but, for them, Northern Ireland is quite a small market. That requires us to have a very close relationship with those who have the technology expertise.

We need technological expertise to roll out a new ERP system, Integr8, but we are not going to need it again after that, so it would not make sense to build that capacity up in-house. However, that means that we have to train a lot of people so that they can use it. When they are trained, they will become incredibly attractive on the private market. There will be a constant challenge to keep people. The way to do that is to make people feel that they have a strong career with a good purpose, supportive training, an encouraging structure and the opportunity to work on these key projects. We are making the steps that I mentioned. We are making really good progress, but — I have to be honest — some of those skills are just in short supply globally, and therefore we have to do the best that we can to train our own and be a really good customer to work with for those who help us.

Mr Gildernew: You said before that developing that professional skills pathway takes time. I am curious to know whether we are actively headhunting others to fill some of the current positions, given the fact that that involves a tail or a lag. Are we actively looking at people who we may want to bring in and those who might have skills for a particular stage of a project that we may no longer need? That is one part of it, but I think that the home-grown option is the best and most sustainable. Given the current trajectory of the framework and the current demand from projects, how long will it be before our pipeline meets the needs of projects?

Mr Gibson: As I sit here, I think that it is unlikely to ever be fully in balance. We will always struggle in some ways to get very specific skill sets. It goes back to that point about why you have to be careful about how many procurement exercises you run at once, because there are a finite number of people.

Mr Gildernew: We will come to that one.

Mr Gibson: When Integr8 is in place in 2028, we will have all the job family structures. That will not necessarily ensure that every one of those posts is filled. There is obviously an affordability question there, but equally, to be honest, in the past 18 months, we have seen a slight change in the market when we go out to compete. We are seeing a bit more interest in the jobs that we are putting out relating to delivery and IT roles, but it is still quite challenging to get people. We often have to run repeat competitions because people are not coming through.

Mr Gildernew: I take your point on reaching 100%, so let us set that aside for a moment. When do you think that you will reach 50%?

Mr Gibson: I am very confident that, with the job families in place and the Integr8 solution in place by 2028, we will be in a position to much more clearly identify any skills gaps. For example, I do not think that I am 50% short at the moment on either my Integr8 programme or my LPS programme. I am quite confident that, although it is very challenging, we have a staffing complement that will allow us to deliver. As others do, we have a strategic delivery partner from the private sector. We have developed really close working with our colleagues in other jurisdictions to tap into their expertise, so it is not always the case that we will have to build it. I am confident that we already have a delivery team under Paul and my other SROs that will be able to deliver, even with those challenges of filling some of the posts.

Mr Gildernew: What is your assessment of the current percentage?

Mr Gibson: It is hard to quantify it, but I would put it at around 80% for the projects that I control. Paul, you are the SRO, so you can speak to Integr8. It is constantly a challenge, and we have key gaps, but, when I meet the senior team, I am supremely confident that we have the right capability. That does not mean that we will not hit some hurdles.

Mr Duffy: From an Integr8 perspective, we certainly do not have a fully resourced team. We have struggled to recruit people into roles. There is another issue that we face a little bit. The Integr8 programme gives our staff great opportunities and exposes them to lots of experience in particular areas and ways of working, and that makes them very attractive when it comes to other roles. It is not only recruitment but retention that is quite challenging. It is about trying to retain those staff. If you had job families and career progression, and if people could see a clear career pathway for themselves, that would help with the retention issue. At the moment, we are sitting with 15% to 20% vacancies on the Intergr8 programme at any one time, but that will move as we recruit and lose people.

Mr Gildernew: The delay in some major projects, including IT, feels greater than a 20% deficit, although I am not in a position to answer that.

Mr Gibson: Intergr8 is a good case study. Some of those delays were because I could not secure the funding for the programme — absence of an Executive. You have to bid for it and get security of that funding. That is just one example. COVID would be another for other IT projects in other areas where things had to be put on hold for something that would have been —. It is not always just staffing or skilling. The other common issue that we have run into, certainly in some of our LPS work, is uncovering new things. As you start going out to procurement, you realise that maybe the data is not quite in the shape that you would like. That takes you down a little bit of data cleansing work or something that maybe takes you a little bit longer. Each project has its own story to tell.

Mr Gildernew: I will go back to the number of projects versus the amount of capacity. Are we trying to do too many things at once? Should we be prioritising and putting the skills that we have into key projects that would then allow the others to progress?

Mr Gibson: That is certainly the decision that I have taken as an accounting officer. I have had to be very particular about what I feel the Department can progress. In some cases, that has meant me as a leader saying, "I don't feel confident that we can run this many projects concurrently. There's only so much bandwidth for me and for our board". I cannot speak for others, but that is the way that I have approached it, which is to not run so many of these simultaneously. Integr8 will change completely the way finance teams across all Departments work. It will change our whole HR system. You would not want to bring another systems change to either of those groups at the same time, so we would not. That can often mean me saying, "Well, I'll push that back a little bit".

For example, we have had areas in pensions and elsewhere where you get to the point where you do not have people bidding. People think that it is just not a valuable enough project for them to make a significant return on. You then have to start again, re-spec and have another look at what you can do. You are quite right that that is exactly what has to be done, and, unfortunately, when you get aggregate reports, they will summarise the number that are —. However, the answer would not be to run all of them concurrently; absolutely not. There would not be the bandwidth at senior leadership, resources would be spread too thinly, and you would not be able to finance them either.

Mr Gildernew: In addition to the skills needed to deliver major projects successfully, there is a need for digital skills to deliver a more responsive public service generally. Has any work been done to determine whether those skills are in place across the Civil Service? Also, what is being done to ensure that there is capability to drive forward the use of these modern, innovative digital technologies that will assist us to do things quicker and better, we hope?

Mr Gibson: One of the other areas that I am responsible for and have an SRO for is the people strategy, and digital skills is a component of that. In that strategy, we have dedicated permanent secretaries to look after different components. Ian Snowden in Economy, which you can understand because of the overlap, is looking after the digital skills stream in the strategy. The chief digital officer is one part of that. It comes back to what will be possible in Integr8 that will allow us to be better able to say to staff at all grades and levels what type of digital skill and competency we might expect them to have at a point in the future and how we can provide the appropriate training.

We have been rolling out training for things such as Teams usage and Copilot. Once we have a much better model, however, we will be able to, for example, record on the system what type of training people have done. That will allow us to better identify where digital upskilling is most beneficial. The chief digital officer is probably going to look at — I am thinking ahead; the person is not in post — critical roles across the Civil Service that will have to have almost a digital-first mentality, who think of a digital solution first when they come to a problem. As the person with responsibility for HR, it is important for me to say that we keep all staff on board with this and engage with the trade unions on the shape and nature of the way that jobs are changing. They will all have a digital component going forward. That is why digital skills is one of the themes in the people strategy.

Mr Duffy: We carried out a digital maturity assessment across all Departments looking at current skills. A set of recommendations and actions are being actively implemented on how we upskill people, create greater capacity and grow the digital profession. That work has already started.

Mr Burrows: Thank you.

The Chairperson (Mr McCrossan): Is it fair to say, Neil, that the development and implementation of the IT systems is a form of transformation?

Mr Gibson: Yes.

The Chairperson (Mr McCrossan): In the headlines all week, there has been huge concern about the amount of unspent transformation funding. That money was allocated by your Department, but was there nothing specific in the allocation for transformation in your Department?

Mr Gibson: There was nothing in the first phase. We have projects in the second phase, which has just been announced, including our digital workplace project. Under the agreement with the UKG, there was a small amount in the transformation fund for support for delivery. The board got agreement to utilise some of that to fund the hiring of a chief digital officer so that there is a digital lens on the transformation board. The digital workplace project, which is a major digital transformation project for us, is part of the second phase, so we look forward to getting the first of those projects.

The Chairperson (Mr McCrossan): I posed that question because you referred on a number of occasions to having to bid for funding for transformation. Will that situation resolve itself now?

Mr Gibson: Yes. The Department is delighted to have been successful with a couple of bids from the next tranche. That is great. The digital workplace bid is important because of things such as Content Manager and the way in which we store all the data that we hold, which is not in a format that would be appropriate for its interrogation. That will therefore be an important project for us.

The Chairperson (Mr McCrossan): Great.

Mr Dunne: Thank you, gentlemen. I am alarmed about the major capital projects of recent years, because similar themes have come through from them about capacity, capability and, ultimately, the need to improve efficiency. We do not want you to be back here in the months ahead covering issues that have already come up time and again.

With that in mind, I will ask about the Integr8 programme. I see that you have the Integr8 lanyard on, Paul. Integr8 has, as you have said, been a long time in the making and has existed in various guises over the past decade. I am not sure whether you view it as a silver bullet, but I am sure that we will come on to that.

The Account NI contract expired in 2018, while the HR Connect contract expired in 2021. You touched on some of the main reasons that it has taken your Department so long to replace the outdated systems, but I am keen to know how lessons can be learned from that. Are you confident that the estimated cost of £294 million ultimately represents value for money? We are talking about serious numbers — in the hundreds of millions of pounds for Integra8 and over £5 billion for all major IT projects — which represent huge sums for everybody in Northern Ireland and highlight the seriousness of the issues.

Mr Duffy: The Integr8 programme has had a stop-start history, and there are a number of reasons for that. I arrived in the Department in May 2020 to take on the role of looking after the Integr8 programme. That was when we were in lockdown, so, unfortunately, a lot of the resources that had been allocated to the programme were diverted to respond to COVID.

Mr Dunne: I appreciate that.

Mr Duffy: There was therefore a bit of a delayed start. When we got through COVID and started to stand up the programme team again, as Neil mentioned, in 2023-24, we had a year of not being funded. Unfortunately, that led to the programme's being stood down and the loss of almost all the staff, so we then had to go through a process of restarting it.

The programme is now well and truly under way. The business case that was approved in 2023 estimated an overall cost of £294 million to design, build, implement and run the service for all nine Departments until 2034. I am fairly confident that we will not exceed that £294 million estimate. The one thing that I will say as a caveat is that it is difficult, from where we were sitting in 2023 — we touched on this earlier — to write an outline business case for approval of a system that we have not purchased, designed or implemented. A lot of estimates were made at that time, but I think that we have done enough due diligence and pre-market engagement to say that the £294 million figure is fairly robust.

Does it represent value for money? The Integr8 programme went through a proper, robust business case process and was assessed and deemed to provide value for money. There are similar programmes in other sectors, such as health and education. We have also looked at other jurisdictions and benchmarked against them as best we can. It is not always possible to do a like-for-like comparison, but we have a fair indication of how much a similar programme in a similar jurisdiction costs, allowing for differences in complexity and local circumstances. The £294 million therefore seems to be the most reasonable estimate at this stage.

Mr Gibson: It is also worth saying that, for a project such as Integr8, defining benefits can be more difficult than you might imagine. There are obvious savings, such as time savings, but there is also all the stuff that we have talked about: access to better data; understanding our staff better; and being able to help them more to make their day-to-day work easier. That includes everything as simple as booking leave and doing expenses. It is about modernising all of that. It is very hard to say, "That will save three minutes". The quantified benefits are therefore a very conservative estimate of what Integr8 will do, particularly because it should lead to a big cultural change. For example, when people are sitting at a NICS board or programme board meeting, their management information and intelligence about their staff should be at their fingertips, as opposed to an accounting officer having to request it. There are potentially a lot of benefits from that system. We were very lucky to have the Finance Committee out with our Integr8 team yesterday to see some of the work that is being done, and I am really excited about some of the benefits. Those benefits could not be put into the benefits realisation case, because they cannot be quantified, but they should bring about some quite significant culture changes.

Mr Dunne: Can you briefly refresh my memory on some of the key dates for when Integr8's impact will be felt? Is there a timeline?

Mr Duffy: For going live? We have learnt lessons from other jurisdictions from schemes that have not been quite as successful when they went live. We found that, to mitigate that risk, we should try to avoid taking a Big Bang approach, whereby the entire system goes live all at once. We have designed implementation that is to be done on a wave basis, and we will start implementation time next year. Between April and August, there will be three waves, with Departments going live for finance. We have split that across the three Departments. Once finance is live, HR will go live in two waves, with everything to be live in February 2028.

Mr Dunne: OK. That is not that far away. As you say, the system is being implemented across all Departments, yet we frequently hear how DOF has no remit outside its own Department. We have heard that again today. We are keen to establish how your Department will ensure buy-in and readiness across the Departments. What is your assessment of the current state of readiness?

Mr Gibson: That is really important, because Integr8 is an example of a programme that is a shared service, so we do have a responsibility for other Departments there.

Mr Dunne: I appreciate that there is that conflict, in that it is happening in the Department of Finance but also across all other Departments.

Mr Gibson: There has been a big programme, which Paul will speak to, of engaging with all finance directors, involving all the staff who will be impacted on and our trade union colleagues, and making sure that we have had full co-design workshops to determine what they would like to see in the programme and what different groups want to be able to do with the new system. That is one good example of something that is not silo working. I emphasise that strongly. We have talked about the social licence before. It is almost having a social licence across the service for people to embrace the change. How I know that that is working is that I am not getting complaints at accounting officer level or permanent secretary level about what is a programme of disruption. People are instead talking about the benefits that it can bring to them, and that really speaks to the co-design approach that we have been taking. Integr8 is a programme that we are delivering for others, so DOF absolutely has responsibility for it.

Mr Dunne: It is fair to say that you are very confident about the benefits that Integr8 will bring.

Mr Duffy: We are confident about the known benefits. I think that greater benefits will emerge as the system becomes established. With a lot of this, the benefits will come from data, and, as we enrich that data over time, the benefits will increase greatly. For example, a lot of our budgeting and forecasting at the moment is done off-system. It is all done on spreadsheets. There is a huge amount of work required to check and maintain accuracy, produce those spreadsheets and summarise them.

There are therefore limitations to that. There are risks around accuracy and issues with timeliness, by which I mean how long it takes to get the data. Working off-system also limits staff's ability to do any sort of analysis or run what-if scenarios. Having a lot of that information on the system, at the touch of a button, will bring huge benefits, and I think that those benefits will materialise as we start to use the new solution.

Mr Dunne: Thank you, gentlemen.

Miss Dolan: I thank Paul and Neil for coming to the Committee. I am sorry that I cannot be there in person, and I am sorry that I missed the Finance Committee trip yesterday. It was one of the few Finance Committee meetings that I had been looking forward to.

The NIAO report states that Integr8's "benefits have not been quantified", but, given your answers to Stephen, you seem to be confident that the benefits will justify the cost.

Mr Gibson: Yes. There is a benefits realisation plan, and we have some quantification of time saved etc, but, as I have reflected, there are a lot of benefits that are much more difficult to quantify. I also worry about the amount of work and effort that one might spend on trying to measure the benefits, when a lot of them will come through in the way in which people behave differently and the time that it saves them on the most routine of tasks.

This will sound like a very philosophical point, but if something is easier and quicker to do, people's reduced levels of frustration will be quite important for their mood and performance. There are therefore a lot more benefits beyond the obvious saving of time or in paying suppliers. That is just one part of that. Integr8 should free up some talent across the service to do other activities and to do some more value-added work with the data, rather than their doing the processing jobs that Paul spoke about.

Miss Dolan: That sounds quite positive. Will you take us through how the business case was approved if the cost was not really justified at the time?

Mr Duffy: We took a two-stage approach to the business case. The final business case, which captured the overall cost and the financial and qualitative benefits, was assessed by economists as representing value for money and was therefore approved. We have estimated that, when the programme goes live and the system stabilises, the financial benefits will be amount to around £14 million a year. That is based on the fact that, at the moment, around 1,200 staff in the finance and HR professions are working in those functional areas. If you look at the size, shape and complexity of the Civil Service and benchmark that against industry standards, you will determine what the size of the finance and HR functions should be. The difference between what we have now and what we will potentially have in the future gives us an indication of the potential financial benefits from releasing resources.

We have set out in the business case a range of qualitative benefits that will come, including an easier-to-use system for staff so that they can access information better and process things quicker. Automation and data analytics have been built in, which means that there is greater reporting and insight to inform financial and workforce planning. Another one of the benefits that will come from the system — we talked about this earlier — is continuous improvement. When the system goes live, there will be continually updated functionality. That should minimise the future costs of replacing the system, which we would normally have had to prepare for, because new functionality is available. It will be readily available as and when it is released by the Oracle solution that we will be using.

Mr Gibson: Sometimes, when annual savings are quoted, one thinks, "How many years does it take to make them?". That is the wrong calculation to use, because the service is paying every year for a set of services. People will be getting paid, and staff will be able to do their invoicing and all their financing. That is an actual service that it is delivering, which brings with it the according value.

Miss Dolan: Is the system therefore future-proofed? Is that what you are getting at?

Mr Gibson: It is, to the extent that one can ever say that, in the sense that it will be constantly updated and will have a lot of potential to develop. We cannot, however, say with confidence that anything is future-proofed. In addition, to exploit Integr8 to its fullest, we will need our staff to be able to use it fully, so the big benefits that we will see, and the ability for it to be future-proofed to an extent, is our being able to think, "How can I use this in a new way? In what other ways can I design tools that I need?", as well as to think about things that we may not have even imagined. As best as we can, as Paul has articulated, moving to a situation in which a system is constantly updated — that is really a service on demand — means that we will always be right at the cutting edge of what that supplier is providing. We will therefore be right up there with best in class with the tools that we have at our disposal. We then just have to make sure that we have the surrounding infrastructure to deploy and utilise them fully.

Mr Duffy: To add to that, one of the big lessons learned from the previous Account NI and HR Connect systems is that they were heavily customised to meet Civil Service needs. At the moment, if we want to make a change to those systems, we are heavily reliant on a supplier, and it can be costly to make changes. We are therefore driving standardisation and using almost the out-of-the-box technology for the new solution. Changes can therefore be applied much easier to the solution, which, hopefully, will protect us going forward.

Miss Dolan: Thank you. Talking about going forward, and sticking to a subject that has been raised today, will Integr8 improve the availability, quality and consistency of data needed for effective use of AI?

Mr Duffy: At the moment, we are using what is called an Oracle Fusion solution. It has a limited amount of AI built into it at the moment, but, as I said previously, that will grow over time. At the moment, the AI ability amounts to the use of a chatbot for staff that accesses the data that is in the system, but that data will become readily available to be exploited by AI in the future. It is secure. It is held within what we call a UK Government cage, so the data is not offshore. There is a lot of security and protection as well. Staff will be able to access that data by having visibility that is based on their role in the Civil Service. That means that the data is locked down to what they need to see, and only what they need to see, so such protections are built in.

Mr Gibson: We will most certainly have a better quality of data with the Integr8 system than what we currently have access to about our staff, about their skills profiles and similar material and about the analytics that we can do. Its potential is probably more than any of us can imagine at this point, but it will allow us to take a much more standardised approach to data. As I mentioned earlier, a big cultural and behavioural change for us across the system is increasingly to bring our ways of working alongside how others work, rather than to look for a system that meets our special or specific needs. That means spending time with staff in order to understand why they do what they do, to understand what benefit that brings them and to help them understand how a more standardised solution can give them tools and insights that they had not thought of before.

I am a man who loves a dashboard and a spreadsheet. I am particularly excited about the level of analytics and intelligence that departmental accounting officers and the NICS board will be able to get on the sorts of things that we want to access at the moment but cannot. Like all things digital, the system is judged on the quality of its data, so a big part of the work that we are doing at the minute is work to get that data into shape to analyse it.

Miss Dolan: Speaking of data, this Committee and all other Committees frequently hear that data-sharing difficulties are as a result of GDPR. Are there therefore requirements across new IT systems and platforms to enable the sharing of data? Have any legislative barriers been factored in?

Mr Gibson: There are areas, not ones that I would necessarily know offhand, in which there have been a number of challenges with data sharing that are particularly relevant to us. NISRA is one such example, as other types of personal data is being accessed. We have to be careful there, because staffing and personnel data is involved, but we have navigated any data-sharing agreements that we need for Integr8. For Integr8, that is not really a huge issue. The system is not public-facing in that sense, but you are right to draw attention to the fact that data sharing can come under different legislative remits.

I have been involved with NISRA to get access to PAYE data to help with the NISRA statistics. There are huge amounts of legislative and security processes to go through to make sure that it is bespoke for each individual. On our Integr8 side, however, we are confident that we are OK on that front.

Miss Dolan: Thank you. That is great. I have one final question, which was touched on earlier. In taking an adopt-not-adapt approach, how will changes and costs be controlled? Can Departments still request modifications? If so, will there then be an impact on consistency?

Mr Duffy: My answer, Jemma, if Departments are listening, is that, no, we will not tolerate any modifications

, largely because of what I tried to set out earlier, which is that the real efficiency comes from standardisation. The more that we standardise, use out-of-the-box technology, start customising and trying to say that we are different from other jurisdictions and other areas, the more that we will achieve the greatest benefit. With the Oracle solution that we have purchased, the finance and HR processes are all based on best practice. We would therefore need to have a very strong argument for moving away from best practice. I would almost wear a badge promoting the adopt-not-adapt principle when I talk to people, because that is really important if we want to make Integr8 the most cost-effective, efficient service in the public sector.

Mr Gibson: You cheered Paul up by asking him that. [Laughter.]

Miss Dolan: I do not know what he means by saying "if" Departments are watching, because of course they are watching. Does not everybody watch PAC meetings? Thank you very much for those answers.

The Chairperson (Mr McCrossan): Neil, you and Paul will be glad to hear that that is the end of our questions. I thank you both for your attendance and for the evidence that you have provided in response to our questions, some of which were uncomfortable and some of which were very comfortable, including that last one, Paul.

It is important for those who are listening in that I reiterate that, time and time again, the Committee has identified the same underlying problems: strategies that are unrealistic and, in some cases, simply undeliverable, or, in this case, non-existent or at the point of approval; a lack of governance and accountability; insufficient capacity and capability to deliver complex change; and an over-reliance on extending legacy systems when delivery falls behind. Those are not in any way new issues. We are well versed in knowing the challenges. They are persistent failures that continue to expose public services and the public purse to unnecessary risk and significant costs. Fit-for-purpose IT is now fundamental not only to public services but to our very way of life. When it goes wrong, the impact is felt directly by the citizens who are affected and by our front-line staff on the ground. When it is done properly, however, it can transform services and improve outcomes and, in turn, people's lives. The Public Accounts Committee expects to see clear evidence that lessons are being learned and that future investment in, and delivery of, major IT projects genuinely supports effective public service delivery.

We have been quite robust in our questioning of you both today. I appreciate that, at times, questions cannot be answered in the way in which the Committee would like or, in some instances, would expect. We do, however, appreciate the time that you have given us today. I am sure that you appreciate the deep frustration that is felt about that particular issue and the potential that it has to affect the daily lives of our citizens.

Have you anything that you would like to say, Stuart? If so, join us at the table.

Mr Stuart Stevenson (Department of Finance): I have just a contextual point to make, Chair. I will build on the line of questioning from Mr Gildernew earlier about capacity and capability. I am thinking in particular about the Integr8 project. He talked about the importance of the skills that are required in the IT profession, about project management and about SRO skills. The Committee should not lose sight of the impact that the project will have on the finance profession. I am speaking for the profession collectively here, Chair, about its key outputs: its statutory requirement to produce accounts; and its role in budget management, cash flow and supporting the audit process. The scale of the Integr8 project is putting extreme pressure on the finance profession. We are aware of that pressure, and the risks have been reported. We have built into the project things such as staff substitution budgets to help mitigate those risks. I want to reassure the Committee that such things are in place, but, nonetheless, significant intellectual property from the finance profession is being pulled into the project. All our outputs are central to the work of the Committee, so we will be doing our very best, collectively as a profession, to manage the risks over the coming years. We are so much focused today on the Interg8 project, so that is an issue of which the Committee should be mindful.

Mr Wilson: The word "complexity" has been mentioned a lot, and I am curious. This question is for whomever wants to answer it. How do you see the expertise, within and without, being stacked? It is a hugely lucrative business for outside agencies to come in and provide IT systems. I am worried that, if that level of complexity is involved, it will be easy for the Civil Service to hand over authority to someone with greater knowledge of IT systems. What will be the ultimate cost of doing that. How will we then be able to cut through that and say, "The industry is wrong, because our civil servants have knowledge of aspect a, b and c and have proven that that is not correct and that here is a better system"? How much complexity is involved? Are we over a barrel to an extent, and, if so, how do you mitigate that? That is key to ensuring that we not sold a pup and thus avoid a situation in which technology that you have been told will last for x number of years is all too soon proven to be ineffective, which would result in our having to spend again to find another solution.

Mr Gibson: It is very important that our procurement is run correctly to help test that. It is also important that we rely on external expertise to help us assess such things.

Mr Wilson: Is it consultative?

Mr Gibson: Yes, we have a programme board for Interg8 that has experts on it from GB and elsewhere.

Mr Duffy: There are a number of ways in which we can try to mitigate that risk. As Neil mentioned, I have two critical friends who sit on my programme board from the UK Government. They are not consultants but civil servants who have been on a similar ERP journey in a large Whitehall Department. They offer challenge and support to me on how the programme is being delivered, and they give me advice. When a supplier presents to me, I often ask for their view. If something is proposed to us, we also do a lot of benchmarking against costs. We engage widely with other jurisdictions. The equip programme in the Department of Health is very similar. It is installing an Oracle solution as well, so we have monthly meetings with it to check what its supplier is telling it, what the costs are and what issues are arising. Likewise, we do that with the Scottish Government. We are therefore doing lots of outreach in order to understand and to be as informed as best we can. Using Oracle is a very special and technical skill, and that skill is not readily available in the Civil Service. We therefore have a separate contract to bring in some individuals with that expertise by way of staff substitution, so they are embedded in our team and face off against suppliers. We have that expertise in the team to give us independent advice on what a supplier may be telling us. We therefore have a number of mitigations in place to try to ensure that we are the most intelligent customer that we can be.

The Chairperson (Mr McCrossan): Thanks, Paul, Neil and, of course, Stuart. Dorinnia, do you have anything that you wish to add.

Ms Dorinnia Carville (Northern Ireland Audit Office): No.

The Chairperson (Mr McCrossan): You are happy enough. OK. Neil, is there anything that you wish to add before we finish?

Mr Gibson: No, other than to say thank you, Chair and Committee, as always. These sessions are important for us as well as we continue to drive improvement. Thank you very much.

The Chairperson (Mr McCrossan): Thank you. Thank you for attending and taking our questions.