Transformation loading: leaders share insights on unlocking the power of data and AI in government
If governments are to modernise and meet public demands, they must embrace data and AI to make efficiencies, produce more informed public policy and deliver improved services for citizens. So, what’s standing in the way? And what can success stories teach us about the road to future government?
Data and artificial intelligence is increasingly central to how governments make decisions and to enhancing service delivery, policymaking and operational efficiency. While strong progress is visible in some areas, systemic barriers need to be addressed to enable larger-scale transformation.
During a Global Government Forum webinar, supported by knowledge partner SAS, a panel of public and private sector leaders from France, Spain, Canada, the US, and Brazil discussed how governments can coordinate AI and data implementation through the development of interoperability and data standards, and how best to share and learn from successful AI use cases.
Seong Ju Park, policy lead on digital government and data within the Organisation for Economic Co-operation and Development’s (OECD) Digital Government Unit shared some of the AI trends she is seeing across OECD member states.
She said governments around are the world are experimenting with and using AI in a range of areas and for many different tasks – including service design and delivery, citizen engagement, document processing, forecasting, risk detection and fraud analysis.
However, she highlighted that a growing number of use cases does not mean AI is either “mature or scalable”, and that the countries making visible progress are not necessarily those with the most AI pilots.
Scaling AI, she emphasised, depends heavily on the quality and accessibility of the underlying data that powers it, as well as other factors that create the right foundations and conditions for success.
“Practices [involving AI] are more established where administrative routines already exist, such as in policymaking, planning, evaluation and regulatory oversight. However, data-driven approaches to service design and delivery and cross-sectoral policy initiatives tend to be uneven across the OECD membership because data is still often fragmented across institutions or not collected in a way [that enables it to be] shared and reused,” she said.
“There are [examples of countries] that have invested in the conditions that allow experimentation to translate into operational impact, by taking into consideration data quality, interoperability, sharing arrangements, and stewardship of data.”
The problem of decentralised data
Jennifer Robinson, global public sector strategic advisor at SAS, agreed with Seong Ju’s point about the extent to which governments struggle with their data.
“They need to be able to get their data together to be able to actually do AI,” she said.
She referred to a report SAS produced with research insights from the global tech intelligence company IDC, in which more than half of respondents cited decentralised data as the biggest impediment to being able to use AI.
Interestingly enough, Robinson noted, about half of government respondents also reported that their data is standardised.
This means “government data is often defined and structured and formatted in a way that allows the administration to have ‘one version of truth’ across their organisation, but only about a quarter of governments have their data managed, and this is becoming a real challenge for governments that want to use AI”.
Echoing Seong Ju’s point, Robinson also emphasised that the proliferation of AI pilots does not necessarily produce AI maturity.
“We’re seeing that the public sector is lagging the private sector in its execution of operationalised systems. We think in large part that is probably because they don’t have their data optimised,” she said.
Using data and AI to aid government’s support for AI innovation across Canada
Kim DesLauriers, the director of the AI Compute Access Fund, which sits within Canada’s Department of Innovation, Science and Economic Development, explained how those working on the fund are taking steps to accelerate AI innovation and commercialisation for companies in Canada, and how data and AI is being used as part of the programme to deliver client services.
The C$300m (US$211m) fund – which was launched last year – has received proposals from all parts of the country and across a wide range of sectors and showed “that there is a thriving ecosystem” of AI innovation in Canada, she said.
“We spent last year learning a lot from our clients about what proper AI innovation research and development cycles can look like.”
The competitive process was launched online and every field of the application form is “delivered to us via a digital path meaning that we’ve really been able to lay the groundwork this year to collect all of the information from innovative companies and proposals for government funding”.
The assessment of proposals is human-led but the system helps ensure each is scored in a specific and standardised way.
“We’ve been able to collect tremendous amounts of data that will then be used to train AI models… to help our public service officers [provide] more streamlined services, faster responses and more consistent evaluations” of proposals received across the Department of Innovation, Science and Economic Development, DesLauriers said.
Read more: Lifting the limitations: a deep dive into enabling data-driven decision-making in government
Testing innovative high-risk AI systems in the EU – and Spain’s health data platform
Francisco Javier Torres Gella is secretary general of the Spanish AI Market Surveillance Authority (AESIA), a national agency that develops market surveillance and supervises high-risk systems, coordinates international engagement, disseminates relevant information, and has established a “think-do tank” to anticipate future technological trends and help it propose new regulation strategies.
It also develops sandboxes for “controlled testing” of innovative, high risk AI systems in the European Union.
He shared details of a pilot sandbox the organisation launched in 2023 within the country’s Ministry of Digital Transformation.
This, Torres Gella explained, had helped the authority “learn a lot” about developing AI systems within the confines of the European Union AI Act, and demonstrated that the more participating companies planned, the more efficient and successful their AI innovation was likely to be.
It also showed that bureaucratic barriers significantly slowed project timeframes in large companies compared to startups, which can be more agile.
“We have seen possible organisational improvements to overcome barriers and advance models,” he said, adding that the sandbox had also enabled AESIA to compile and publish evidence-based technical guides on the development of AI systems by enterprise type and in accordance with the EU’s AI Act.
Torres Gella moved on to telling the webinar’s audience about a national health data platform AESIA launched in January this year.
This was built to enable researchers to access the secure, high quality data that is “essential for developing, testing, and deploying AI tools that can support better diagnoses and more effective treatments”.
He explained: “Any Spanish researcher can access health data existing in Spain. This data space brings a common framework to improving interoperability while allowing each administration to retain control over its own data.”
Why ‘data alone is not enough’ to achieve transformation – and Mato Grosso’s public policy evaluation portal
Sandro Brandão is sub-secretary of planning and digital government for the planning and management secretary of Mato Grosso state in Brazil.
“Data governance, digital transformation, technology and innovation are not isolated initiatives but central elements of a long-term developing strategy,” he said.
His role is to connect “these dimensions” and work “to build a more intelligent and adaptive” state through better and more coordinated digital infrastructure that can help improve government performance and inform public policy decisions.
He stated that while Mato Grosso has “a lot of data” that isn’t enough to transform government and what’s important is enabling the ability to learn from data.
Part of this effort is the development of Avalia gov.mt, the Government of Mato Grosso’s public policy evaluation portal, which was developed to promote transparency, facilitate access to information and support citizen participation in the evaluation and improvement of public policies.
It “integrates planning, execution and evaluation data” and its main objective is to “improve how government data is used to support decision making”, Brandão said.
The impact of such projects is primarily cultural, he added.
“We can now support leaders to understand what is working, what’s not, and what should change… Data is no longer [just] something we publish – it’s something we continuously learn from”, including to inform how to better govern the use of AI.
The data tsunami, fears over using AI responsibly and other challenges
SAS’s Jennifer Robinson also discussed the ways governments are working to make their data more usable, shareable and effective.
“One thing we are seeing, which is extremely exciting, is that we are moving into being able to analyse unstructured data, and when I talk about unstructured data, I mean long-form text, video, images and audio.”
For a long time, organisations were limited to analysing data only in tables, she said, but with the advent of technologies like natural language processing and large language models, that is now changing.
However, she highlighted that this development also represents a new challenge for governments, because the volume and variety of data now available to be analysed is so much greater. This may prove “both an enabler and an impediment… as governments try to get their data consolidated and optimised for analysis”.
Another challenge is civil and public servants’ hesitancy to embrace data and AI for fear of causing a security or privacy breach.
“I was once guilty of that hesitancy, until I started making this deliberate shift to understand all of the requirements that are put in place to make sure that we have safe systems for our citizens,” Canada’s DesLauriers said.
AI systems might at first appear daunting, but on closer inspection there is usually a “common sense and practical approach” to deploying them ethically and safely.
“It’s not as daunting once you make that deliberate shift [to understanding the guardrails],” she said. “One thing I’ve really been focused on is [telling] my counterparts and other leaders across the public sector [that] it’s not so complicated.”
In closing, Robinson said that the principles of data and AI responsibility boiled down to “human centricity, inclusivity, accountability, privacy and security, and the robustness of data”.
However, while governments may broadly understand and accept these principles, the question of how they ensure these are built-in and adhered to remains.
“That’s the challenge for governments right now, and that involves oversight, tending to compliance, culture, your operations [and] using tools that embed the principles of trustworthiness into them,” she concluded.
The ‘Unlocking the power of data and AI in government’ webinar was hosted by Global Government Forum with support from knowledge partner SAS, and took place on 7 May 2026. You can watch the webinar on demand here.
About Jack Aldane
Jack is a British journalist, cartoonist and podcaster. He graduated from Heythrop College London in 2009 with a BA in philosophy, before living and working in China for three years as a freelance reporter. After training in financial journalism at City University from 2013 to 2014, Jack worked at Bloomberg and Thomson Reuters before moving into editing magazines on global trade and development finance. Shortly after editing opinion writing for UnHerd, he joined the independent think tank ResPublica, where he led a media campaign to change the health and safety requirements around asbestos in UK public buildings. As host and producer of The Booking Club podcast – a conversation series featuring prominent authors and commentators at their favourite restaurants – Jack continues to engage today’s most distinguished thinkers on the biggest problems pertaining to ideology and power in the 21st century. He joined Global Government Forum as its Senior Staff Writer and Community Co-ordinator in 2021.
Related Posts
