Maximising the potential of government data: insights from GGF’s Data Summit
Data is fundamental to the work of modern government.
To support civil servants in maximising the value of the data they collect, and in harnessing its potential to drive developments in areas such as artificial intelligence, Global Government Forum convenes the annual Data Summit to bring together data leaders from across the world for a series of roundtable discussions.
These discussions are designed to foster open, in-depth dialogue on overcoming key challenges and sharing global best practice.
The summit provides a safe space for a candid and open conversation about how to improve the vital data infrastructure that government runs on.
To coincide with the publication of the agenda for the 2026 Data Summit, this article provides a summary of the key issues discussed in the 2025 edition:
- Paving the way for AI with data governance
- Making sure government meets its priorities: how government can use data to measure performance
- Data sovereignty: why it matters and what to do about it
Developed in partnership with public servants internationally, the Data Summit explores how data can power the government of the future, with the forthcoming 2026 edition set to deepen the conversation on data governance for AI, making data sharing happen, understanding the implications of data sovereignty and treating data as a national asset; and building and sustaining digital and data capability and talent in the public sector.
Summit discussions were held under the Chatham House Rule to allow participants to speak freely, so this article provides an unattributed summary of the discussions across the key topics as a look ahead to the 2026 discussions.
Find out more about the Data Summit 2026, being held in London on 14 October 2026
How governments can pave the way for AI with data governance
This discussion focused on how ready data in government is for use in AI – and the practical steps governments can take now, and over the longer term, to ensure they can capitalise on the benefits of AI. The session, supported by knowledge partner NTT Data, a leading consulting and IT services provider, also covered the complexities of AI and legislation.
Speakers highlighted the need for data to be used to help ensure AI works for everybody by reflecting “the values that we stand for, and the standards that we expect”, one leader said.
They spoke about the work their government had done to both define core standards to unlock services and AI innovation, and develop guidance for how to use AI to identify insights in unstructured data.
“We must ensure that ethics is not seen as a brake on innovation, but instead seen as a steering wheel,” they said.
Governments therefore need to invest in data governance and data sharing, they added.
“Data governance is frequently misunderstood and under invested, yet it’s a critical enabler for building a future powered by data that you can trust.”
Another leader said central government departments need to identify where to add the greatest value on issues like data standards. There needs to be better conversations in government about how to define minimum, viable data standards to address “quite a lot of inefficiencies and duplication across the wider public sector”.
Investment is “absolutely necessary” to address fragmented datasets in government, one attendee flagged “There’s a lot of duplication going on… We need to bring data together and provide hubs for all relevant authorities to be able to dock into and openly share data, and a lot of our evidence points to the fact that we [in government] are not very good at that.”
However, others highlighted countries that have done this well. They have developed data hubs, for example, which provide access to high value datasets and are focused on providing data for AI training – and they have become more advanced in AI adoption.
“The most successful way of getting investment in data is actually to tie it into the operational imperatives of the department,” one speaker said. “It’s not to have a standalone programme, that never works. You end up boiling the ocean. You end up not actually getting the investment you need, and you don’t drive the change of pace you need.
“I found I was much more successful when I tied [data] into the missions and strategy of the organisations I worked in, and then actually orchestrated that.”
GGF’s Public Service Data.AI conference, focused on building smarter public services through data and AI will be held at the Business Design Centre, London on 15 October 2026. Find out more and register here.
Moving past technical language
Concerns were also raised that, in the words of one attendee, data leaders are “sometimes a little bit guilty of being technical in some of our language”. But, they added, it isn’t technical leaders, in most cases, who hold the purse strings but rather other business and finance colleagues. “What I’m trying to do with my team is understand the language of the business so that we can land the kind of funding that we need. And that is really thinking about articulating the outcomes [and] what will be different.”
Data governance can be part of the toolkit that helps governments navigate it.
“We’re not waiting for AI to arrive at scale. We’re preparing government now to use it wisely, ethically and effectively,” one leader said. “It’s about creating a public sector where data is not a barrier, but a bridge to better outcomes. With the right data governance in place, AI can help us to cross that bridge – safely, responsibly and with purpose.”
Another attendee emphasised that governments “need to invest in business and culture change to help people work alongside AI”.
They added: “We need to incentivise people to learn and adapt to the new technology and to learn to collaborate with it. We need to be able to optimise processes with AI.
“I’m talking about all the business change that sits around the AI solutions, as well as the data governance foundations.”
This includes greater data literacy as part of this process. “The civil service needs to invest in colleagues who really understand the issues around AI and what the capabilities are: from defining good requirements for AI solutions through to understanding the quality of AI products.”
Other sessions touched on similar points, highlighting the need to tackle what was called “sludge” in government.
“I would say that the bureaucracy of government [has] almost got to the point where it’s very hard to make decisions and take action, because bureaucracy has got in the way,” one leader said.
“The opportunity we have with innovative solutions is to [say]: I can give you this, but I also need to look at how you work, and I need to help you refine yourselves a bit.”
Read more: UK launches National Data Library with early years kickstarter project
Using data to measure government performance
The data that governments around the world collect provides vital insight into the services they deliver, and for public servants it is imperative that they use the information their work generates to measure how well delivery is progressing.
In this session, supported by knowledge partner Deloitte, a leading provider of professional services, the discussion focused on the importance of data sharing, the drivers of government data usage, and the challenges in cross-government data sharing.
Throughout the session, leaders shared examples of data sharing leading to better services, including governments sharing and connecting data across health, social and family agencies – and even across borders – to reduce fraud and address areas of non-take-up of services.
However, the session also examined how a lack of data join up was “restricting the ability of governments to deliver”, and that there is more to be done to ensure data is “not just to measure the past but shape the future”.
The discussion touched on the difficulties in achieving this – but also on how to make it happen.
As one speaker put it, where linked datasets exist to help drive better understanding of services, it can be “incredibly powerful” and help improve services and drive take up.
New Zealand was named as a good example of where citizen satisfaction surveys are used to make “almost immediate improvements to the underperforming service delivery channels, whether digital or in person”.
Other governments that are also good at ensuring high-value data is shared include South Korea, France, Estonia, Slovenia and Canada, speakers highlighted.
However, where the data sharing does not exist, it is restricting the ability of governments to deliver by making it difficult to create the feedback loops that help improve services.
Creating systems to help government use data means embedding data collection and analysis into the regular policy cycle – moving from reacting to problems towards predicting them – and attendees said the biggest driver in making this happen is leadership.
“If [backing] comes from the top, you can unstick things,” one leader said, but “it’s almost a lack of confidence that somehow, something will go wrong”.
“It’s in those countries where there are very bold leaders that are willing to move forward that we are seeing a huge change.”
‘Only share the data you need to’
Others spoke about the processes of linking datasets to help understand the effectiveness of government delivery.
Such linking increases the analytic potential of data, and allows datasets to be analysed on the basis of geographies or industry or occupations.
The aim is to move away from departments asking “how can you share all your data with me?” to instead “how can you share just the bits that I need to do this bit of analysis?”
As one leader said, data owners are “rightly very cautious” about sharing their data, and a need was identified for an “overarching vision of where are we going” to inform a data sharing strategy, “and I don’t think we’ve really got that yet”.
One attendee concluded that data teams need to be “among the best if we want to align in outcomes”.
“We should be among the best at defining what value is, and measuring that value… We should have a coaching capability that delivers governance and helps teach people how to make decisions from the data and the insight that our functions deliver.”
The performance frameworks needed to do this are “extraordinarily difficult to build”, they added – but “they are worth building” to provide definition against a set of outcomes – and measure progress against them.
“Outcome frameworks are about declaring what you want to achieve and then imperfectly measuring them, and then tuning in your framework to make it better.”
To make progress, they concluded that there was a need to skill the data function, improve understanding of the work of the data function, and better evaluate outcomes.
“Those three things aren’t typically seen as kind of raison d’etre of data functions, and I would love that narrative to change.”
It is now “data’s time” in government, they said. “But because it’s our time, that doesn’t mean that the conversations that we have with leadership are any easier. So I think it behoves us to get to the front on outcomes and make the impact that we know we can make.”
Data sovereignty: why it matters and what to do about it
In the age of increasing geopolitical uncertainty, governments around the world are also increasingly grappling with the idea of data sovereignty. This refers to the idea that data – such as intellectual property, financial data, or personal information – collected or stored in a particular geographic location, such as a specific country, should be subject to the laws of that location.
Providing this security is one way that governments can boost trust in how data is used in government, but there is increasing concern that governments do not have full sovereignty over their data when they store it in the cloud.
Some governments have warned that this has the potential to damage trust, and this session highlighted how governments are thinking about boosting national control and security in how data is used.
Speakers in this session highlighted the increasing challenges governments face in balancing security, trust, and economic opportunities.
One speaker said the concept of data sovereignty has become increasingly central to the government, due to concerns about the potential loss of public trust from data being stored in different jurisdictions. “We can lose that trust in government if folks don’t think their data is kept safe [and] it can be hard to rebuild.”
However, it was also acknowledged there is a need to strike a balance in order to ensure that government can keep pace with expectations – and that total sovereignty is likely impossible.
“Cloud and digital platforms are critical to modern service delivery and innovation,” one speaker said. “We must weigh that against the legal and economic realities of sovereignty.”
Governments therefore need to think about the risk appetite around data storage and hosting.
In one government, discussions around risk are being held on a case by case basis, with a central board formed as a way to “funnel projects through and get a line of sight into what departments are doing, to identify risk, and ways to mitigate those”.
Read more: Governments of UK and Canada announce plans to secure AI sovereignty
The conversation moved past risks to the “great opportunity” to use data as a national asset in the public sector.
Governments have a chance to “really start to think about our data and the value that can have” as a sovereign asset – and one government leader highlighted that their country was implementing a system to value government data in national accounts by 2030. “We will have a figure on data as an asset… and this will be a new methodology.”
Others discussed how they were looking at how to raise revenue to help cover the costs of providing data services, a point that was raised by many of the speakers.
One speaker said: “We’re looking at the value of data as a service, what service levels we might offer, and whether there’s a monetary charge that helps us deliver that sustained service.”
This organisation is also looking at how they might be able to understand how they could benefit from avoided costs as a result of data being used for predictive purposes.
Another highlighted the work that the Isle of Man is undertaking to create the world’s first legal and operational framework specifically designed to treat data as a capital asset – thus enabling businesses to recognise, govern, register and unlock the value of their data, including allowing data to be traded or used as collateral for loans.
However, those around the table agreed that more work was needed to develop a framework on how such monetisation policies might work.
“I think we need to do this as a collective, with some oversight from the centre. That would be my ask. Otherwise we run the risk of getting a framework and then all going up and doing something and potentially having unintended consequences.”
One leader put forward another suggestion: “There is no easy way to monetise or understand how much your data is of value, either as a public service or to be sold. But there are some people who are trying this, and I wonder whether there’s a collection of exemplars or test cases we could use to try and come up with some kind of solution.”
The Global Government Data Summit was held in London on 18 September 2025, and was supported by knowledge partners NTT Data and Deloitte. Register to find out more about the Data Summit 2026.
About Richard Johnstone
Related Posts
