This week, we look at how governments are building, and quietly straining, the digital state - from Hungary’s bid to make public services “invisible” to Norway’s ground-up rebuild of its immigration system.

Our main stories this week:

• Hungary makes the case for “invisible” government as its DÁP super-app tops OECD user-centricity rankings

• Kazakhstan approves a sweeping AI action plan to 2029, days after data on an alleged 15 million citizens surfaces for sale

• Norway rebuilds its immigration system from the ground up in a six-year, $185M digital overhaul

• How AI is breaking the British State

As a reminder, we have recently begun to expand our digital government coverage away from just Substack and onto Linkedin, Twitter and Instagram. For original content, audience engagement, and roundups of all our pieces, feel free to follow us over on those channels. Also check out our sister newsletter, interweave.africa.

Building invisibility into digital public services - Sol Gonzalez and Amit Roy Choudhury, GovInsider

The Digital Hungary Agency (DHA) is pursuing a deceptively simple goal, having public services that are so seamlessly integrated citizens stop noticing them - "like using electricity", in the words of Deputy CEO Zsolt Bányai.

• Operating under the Ministry of Science and Technology, the agency has built one of Europe's fastest-growing digital service ecosystems - anchored by the DÁP mobile app

• Hungary now ranks first among OECD countries for user-driven public service development.

• Bányai contrasts this with the old model of experts drafting theories in isolation and passing documents between ministries. He recalls a "3,000-page system blueprint", in which page 174 contradicted page 2,725. The new approach assumes that a mobile-first, user-centric design will make the state presence feel less like a bureaucratic body and more like a form of public infrastructure.

Can Kazakhstan’s AI strategy deliver both growth and smarter governance? - Ayana Birbayeva, The Astana Times

On the 26th of August, Prime Minister Olzhas Bektenov approved a government action plan setting measurable digitalization targets through 2029 under the national Digital Qazaqstan Strategy - embedding AI and real-time data into education, healthcare, transport, finance and industrial management.

• Rather than digitizing services one at a time, the plan wants to build digital systems across entire sectors, so that data and AI become part of how agencies and businesses actually make decisions.

• The ambition sits awkwardly next to a serious trust problem. This month a seller on the dark web offered a 2.7GB dataset allegedly holding 47 million records on around 15 million citizens - including passport details, phone numbers, emails, employers and passwords - claiming it came from a hack of the eGov portal.

• Kazakhstan's Ministry of AI and Digital Development denied any eGov breach, saying its inspection found the files were a patchwork of images and documents rather than an eGov export. But if confirmed, it would be the second mass breach in just over a year, following the exposure of some 16 million records in June 2025.

Our Take: Kazakhstan’s new digital plan is ambitious, and it has achieved impressive results to date in its digitalization efforts. But the story from The Diplomat shows just how vulnerable that progress can be - with 16 million citizens struggling with leaked records.

Norway overhauls immigration system through six-year digital transformation - LuHai Liang, Biometric Update

Norway has begun a six-year program worth 2 billion kroner (~185 million USD) to rebuild its immigration system from the ground up, replacing decades-old platforms with a unified digital infrastructure spanning visas, residency, asylum and enforcement.

• Approved by parliament in December 2024, the ModUlf project (Modernisert Utlendingsforvaltning) runs from 2025 through 2031 and brings four bodies under one umbrella - the Directorate of Immigration, the Ministry of Foreign Affairs, the police and the Immigration Appeals Board.

• Product lead Henrik Hansen frames the core goal as more stable, modern and safe systems, with the promise of shorter waiting times, real-time application tracking and more transparent processes.

• Officials have been notably candid about the human stakes, acknowledging applicant anxiety and stressing the need for input from foreign residents who often feel like just another anonymous number in a machine.

Our Take: What stands out here is how openly officials talk about the human side, acknowledging applicant anxiety and building in user input rather than treating the overhaul as a purely technical migration. For any government sitting on decades-old legacy systems, and most are, ModUlf is worth watching as a blueprint.

How AI is breaking the British State - The Economist

We have covered before on interweave the idea that AI might not solve bottlenecks entirely, just shift them along processes. In the UK, this is starting to materialize, with AI-written citizen objections and appeals flooding government.

• This challenge has been labeled “agentic flooding”, providing challenges for government in areas ranging from tax appeals to welfare claims to parking tickets.

• Already, the backlog in employment tribunals has risen 55% in a year in the UK, and demand for emergency injunctions has risen 100-fold - largely due to AI-fueled claims.

• The UK is particularly vulnerable to this challenge, since its administrative traditions so favor the sort of written submissions that AI is so good at drafting.

GovTech News in Brief

GDS tests AI to tackle government’s mail backlog - Christine Horton, Think Digital

The UK's Government Digital Service is piloting Project Kestrel, which uses data science to help Knowledge and Information Management teams review decades of accumulated government emails, a task GDS says is no longer sustainable to do by hand.

Pro-Russian hackers claim responsibility for major cyberattack on Norway’s public digital services - The Associated Press, ABC News

A pro-Russian group calling itself Server Killers has claimed responsibility for a three-day denial-of-service attack on the Norwegian Digitalization Agency (Digdir), which a spokesperson called the biggest attack on Digdir's systems it has ever experienced.

Singapore tests chatbot to help citizens cut through govt red tape - Malay Mail

Singapore is trialing a government chatbot built to guide citizens through complex rules and applications, part of its wider push to embed AI agents across public-service delivery.

The Bahamas turns to Estonia as digital identity expertise becomes an export - Lu-Hai Liang, Biometric Identity

The Bahamas is drawing on Estonian know-how to build out its own digital identity system, the latest example of Estonia turning its e-government experience into a growing export.

The UK public sector has spent £1.4bn on AI so far in 2026, breaking previous record - Jack Aldane, Global Government Forum

UK public-sector AI spending has already hit a record 1.4 billion pounds in 2026, showing how fast procurement is accelerating even as questions about value, oversight and lock-in stay unresolved. The figure lands amid a wider debate about whether spend is translating into services citizens actually notice.

UAE launches strategic track of its agentic AI government project - Ali Al Hammadi, Gulf News

The UAE has moved its agentic AI project from policy into delivery, with over 100 federal officials attending a workshop to launch the strategic government track. The initiative aims to convert 50 percent of federal operations, services and tasks to agentic AI-driven models within two years, spanning seven areas from policy development to foresight under the “human leads, AI enables” principal.

The Theory Behind the Practice

Central Europe’s AI rollout needs a public error register - Gleb Tsipursky, Visegrad Insight

The piece argues that as Central European governments push AI into public services, they should keep a public register of failures and errors, treating openness about what goes wrong as a condition for trustworthy adoption.

GI reflects: How digital government has evolved over the past decade - Si Ying Thian and Amit Roy Choudhury, GovInsider

Marking ten years of coverage, GovInsider traces how the public sector went from technology laggard to leader, with digital public infrastructure emerging as proof that private-sector competition alone could not deliver the joined-up services citizens now expect.

The virtues of accessible digital government services - Ross Ashman, Public Sector Network

This piece makes the case that accessibility in digital government is about equity and trust rather than mere compliance, requiring inclusive design, multiple access channels and digital-literacy support so services reach people regardless of ability, location or confidence.

Our Take: This is the necessary counterweight to the invisible government pitch. “Seamlessness” only counts if everyone can actually reach the service and this piece makes the case that accessibility is about equity and trust rather than mere compliance.

Over the past decade, Indonesia has laid the groundwork for its digital transformation through a range of policies, including comprehensive frameworks and strategies.

But delivering this vision across thousands of islands and to hundreds of millions of all while tackling a legacy bureaucratic system was used to operating in silos, has not been easy.

Untangling this complexity – and thinking through how digital public infrastructure (DPI) can bind whole-of-government efforts – was the central theme of Lisa Shaumanissa’s presentation to GovMesh participants earlier this year in Singapore.

Bapennas’ Lisa Shaumanissa presenting at GovMesh 4.0.

“Indonesia is in the process of transitioning towards a fully digital government, and we are trying to use DPI as a one tool to combine all the ministries and sectors to support our public services”, she told participants.

The country has developed tools like Satu Data Indonesia, designed to unify fragmented government data systems, and anchored them around principles of data standardization, interoperability, and shared metadata.

But scaling these tools is easier said than done.

Indonesia’s decentralized governance structure has meant that historically provinces, regencies, and cities have historically developed their own systems in silos. This has created huge amounts of waste and duplication among systems.

The Satu Data Indonesia is now working to address this, actively engaging national and local government partners through virtual and in-person outreach to help them align with Satu Data Indonesia’s new standards.

DPI as the integration layer

This new emphasis on aligned standards is not just centralization for its own sake. Shaumanissa went on to explain how DPI acts as the glue binding cross-sectoral ministerial efforts to overcome fragmented systems.

This architecture spans multiple layers - from foundational infrastructure such as connectivity and computing to data governance through Satu Data Indonesia, and application layers including digital identity, digital payments, and data exchange.

One key component is Sistem Penghubung Layanan Pemerintah (SPLP), a government Application Programming Interface (API) layer developed by Ministry of Communication and Digital to connect systems across ministries.

The social assistance program (perlinsos) framework as a DPI use case in Indonesia.

“They developed it as a gateway to connect one system to another across ministries,” Shaumanissa said.

She also highlighted the importance of reducing bureaucratic layers by simplifying inter-ministerial business processes. With trusted systems in place, data exchange would no longer require paperwork or memorandum of understanding (MoUs).

According to Shaumanissa, Indonesia’s DPI strategy is not only about system integration, but also about redesigning services from a citizen-centric perspective.

She described the “life journey” approach, where services were structured around citizens’ life events from birth and immunization to social assistance delivery.

Particularly successful has been the country’s social protection program (perlinsos). Through an integrated portal, citizens can apply for assistance, which is then verified and processed through interconnected data systems.

The system then assesses eligibility, matches applicants with master data across assistance programs, and notifies the government to disburse support through digital payments.

Continuous learning

Despite progress, Indonesia’s DPI journey remained a work in progress, characterized by ongoing experimentation and continuous learning.

“We are still learning by doing. Not everybody is happy,” said Shaumanissa, noting that trust and resistance within government were the biggest hurdles.

She cited data sharing as a continuing sensitive issue. While Indonesia has made progress in data governance, privacy concerns still influence institutional behavior.

“Indonesia has enacted a Personal Data Protection Law, which sometimes makes ministries overly cautious to share data,” noted Shaumanissa.

As a result, data often remains concentrated within certain agencies, limiting its broader use for policymaking and public service delivery.

A similar paradox exists in technology management. Some ministries are reluctant to integrate, believing their existing systems were sufficient.

To address this, Bappenas has adopted a collaborative rather than a top-down approach.

“We always try to engage and hug them and continuously explain the importance of integrated DPI,” she said.

“This journey is still ongoing. But the direction is clear. We are working hard to ensure everyone understands”, she added.

Governments around the world have seen a significant increase in the amount of data that they collect. However, the use of this increased data in policymaking has not always kept pace.

It is this gap that the United Nations Development Programme (UNDP) is seeking to address through a range of initiatives, including efforts to help governments systematically integrate data into public policy making.

UNDP Global Center Singapore’s Acting Head, Carla Gomez-Briones, told GovMesh that “we have more data available now than at any other point in history, yet it is frequently underutilized in public policy”.

She highlighted that around 90 per cent of the world’s data has been created in the past few years, with the surge driven by advances in emerging technologies.

New data sources have also emerged beyond conventional data - including spatial data, sensor data, commercial data, and data generated from everyday digital devices.

“This presents a new challenge and opportunity for governments: how to manage large volumes of data and integrate them into decision making?”, she said.

In this context, she highlighted the Data to Policy Initiative, designed to support policymakers in grasping the fundamentals of data-driven decision-making.

Leveraging a global ecosystem for innovation

Gomez-Briones began her presentation by outlining the role of the UNDP Global Center in Singapore - a hub for technology, innovation, and sustainable development.

UNDP Global Center Singapore’s Acting Head, Carla Gomez-Briones, presenting at GovMesh

Established in partnership with the Government of Singapore, the center works with a wide range of partners in the country, from government ministries to global platforms such as the World Cities Summit and Asia Tech x Singapore (ATX).

Collaborations also extend to emerging areas such as AI readiness assessments as well as digital trust and safety.

Singapore, Gomez-Briones explained, offers a mature ecosystem - from strong institutions and partners to technological capabilities - that could be leveraged to support developing countries in their own development journeys.

She noted that “we see this as a two-way collaboration, harnessing the Republic’s vibrant ecosystem to support our country offices, while bringing lessons and perspectives from across our global network to enrich partnerships here in Singapore”.

The Data to Policy Initiative

One example of this work is the Data to Policy Initiative, developed with the support of the German Federal Ministry for Economic Cooperation and Development (BMZ) and the European Union (EU).

Beyond technical challenges, a key barrier in fully utilizing data in the hands of public sector entities is the misalignment between data and decision-making processes.

Incentives, institutional structures, and workflows do not always support or enable the use of data in policymaking, Gomez-Briones said.

UNDP and its partners have launched the initiative with a view to creating a global knowledge ecosystem to support governments to leverage traditional and emerging data sources effectively, responsibly, and sustainably.

The four pillars of the Data to Policy Initiative

There are four pillars in the initiative.

1. Data to Policy Navigator: A comprehensive guide to help policymakers integrate data into decision-making processes, focusing on embedding data into real policy workflows rather than standalone analysis. It covers areas such as problem framing with data perspectives, data access, exchange, governance, and visualization - including privacy and regulatory considerations.

2. Training: UNDP also provides interactive trainings based on a “whole-of-government” approach and the data-informed policymaking methodology featured on the navigator, bringing together agencies such as statistical offices and finance ministries to align data use across government and data sharing between institutions.

3. Community: Enabling policymakers to connect, share, and learn from global practices. “We are very much focused on solutions and on creating connections between countries by creating a community of practice,” Gomez-Briones added.

4. Projects: Working with policymakers to integrate data into their policy design. Gomez-Briones shared an example from Moldova, which began with a small pilot before scaling up to a national strategy.

Through its four pillars, UNDP aims to create a global knowledge ecosystem, where policymakers can learn and exchange their experience on data-informed decision-making.

This helps to surface actionable insights on how governments can leverage the power of data to prioritize, design, and deliver for people and the planet.

Tackling real-world implementation challenges

During the Q&A session, GovMesh’s participants raised questions about the target users of data-informed policy.

Gomez-Briones explained that UNDP’s presence in around 170 countries allows the organization to work closely with governments towards achieving their national development priorities.

“By engaging ministries, we can work with officials who are able to implement the initiatives,” she said.

However, she underlined that politics would remain a critical factor. Even with strong technical capacity, initiatives may falter without leadership support and understanding.

As a result, UNDP designed its training programmes to target both technical officials and senior leaders, with the balance adjusted depending on the country’s context and readiness.

Gomez-Briones encouraged participants to explore the Data to Policy Initiative online, highlighting opportunities to engage with its tools, contribute insights, and connect with a growing global community.

This week, we look at Andy Burnham's first act as UK Prime Minister, scrapping the £1.8bn digital ID scheme. Also notable was the EU AI Act crossing from paper into enforcement on 2nd of August, as well as India looking to integrate AI into its Digital Public Infrastructure (DPI) approach

Our main stories this week:

• UK’s £1.8bn digital ID scrapped in Burnham’s first day in office

• EU AI Act enforcement phase begins on 2 August

• India moves to make AI its next Digital Public Infrastructure utility

As a reminder, we have recently begun to expand our digital government coverage away from just Substack and onto Linkedin, Twitter and Instagram. For original content, audience engagement, and roundups of all our pieces, feel free to follow us over on those channels. Also check out our sister newsletter, interweave.africa.

Digital ID plan to be scrapped in Burnham’s first major policy pledge - Adam Goldsmith, BBC News

New UK Prime Minister Andy Burnham is scrapping plans for a government-issued digital ID, originally intended for all British adults.

• Government efforts and resources will instead be redirected towards cost-of-living support and everyday priorities.

• Plans for the digital ID scheme had initially been introduced by Sir Keir Starmer last year, where he argued that a mandatory digital ID would modernize the state, clamp down on immigrants working illegally and enable citizens to prove their identity to access key services quickly.

• Before being scrapped, the government had already reduced the scope of the plans, removing the compulsory element and making it voluntary, suggesting digital IDs could eventually let people manage everything from childcare to tax returns on a “one stop” app.

European Commission starts enforcing AI Act rules and new transparency requirements on 2nd of August - European Commission

Two years after the AI Act entered into force, the enforcement phase has arrived.

• From 2 August, the Commission’s AI Office and national authorities has been able to formally investigate and sanction non-compliance, and new transparency rules have begun to bite.

• Chatbots must now disclose that users are talking to AI, deepfakes must be labelled and AI-generated content must carry machine-readable marks. Fines under Article 50 reach €15M or 3% of global turnover and GPAI breaches up to €35M or 7%.

• Over 180 organizations have signed the voluntary Code of Practice on AI-generated content. The heavier high-risk obligations were deferred by the Digital Omnibus to December 2027 and August 2028 but the transparency and enforcement clocks for GPAI were not moved.

AI as the Next Digital Public Infrastructure (DPI) in India - Understand UPSC

India is exploring how the DPI model behind Aadhaar, UPI and Account Aggregator could be extended to AI in order to make compute, models and data accessible as shared public rails rather than leaving AI concentrated among a few large foreign firms.

• The proposal centers on a “Unified Intelligence Interface” - affordable public compute, open-weight foundation models and multilingual systems built on Indian-language datasets.

• The catch is cost and governance, with high compute and energy demands as well as unresolved questions on privacy, bias and accountability that India would need to regulate before opening AI access at country-wide scale.

GovTech News in Brief

Malta launches €15.6M tender for European Digital Identity Wallet - Lu-Hai Liang, Biometric Update

The Malta Digital Innovation Authority has issued a €15.6M tender to build a fully EUDI-compliant wallet, letting citizens prove identity, verify age, store documents, and e-sign across all EU member states.

Conformance framework for EU Digital Identity Wallets arrives with ARF v3.0 - Chris Burt, Biometric Update

The latest version of the EUDI Wallet Architecture and Reference Framework introduces a set of shared, reusable test cases for assessing wallet functionality, a precursor to certification.

GovTech layoffs part of plans to maintain quality of digital services for Singaporeans - Davina Tham, Channel News Asia

In a rare case of Singaporean public-sector layoffs, GovTech has retrenched 93 staff as the agency moves from one-off vendor-delivered projects to owning and running products itself. GovTech said the shift is not driven by AI or cost-cutting (it retained 102 officers), and it is retraining 110 into new roles through apprenticeships.

South Korea launches free AI agent project for all citizens - Asia Today

South Korea has begun accepting proposals for a government-backed service giving every citizen free access to an AI chatbot and public-service assistant aimed at closing the digital divide. The Ministry of Science and ICT will select two or three private firms with direct-to-public experience, with applications open until 11 August.

The Theory Behind the Practice

The many shades of sovereignty, digital sovereignty in practice - Global Government Forum

Digital sovereignty has moved from an abstract talking point to concrete legal, policy and procurement instruments. This piece argues sovereignty is best understood not as a binary but as a spectrum of choices about control, dependency, and resilience across the tech stack, a framing we aimed to capture in this weeks summary.

Making Agentic AI work for government: a readiness framework - World Economic Forum

The WEF and Capgemini's framework scores agentic AI readiness across 70 government functions rather than ministry-by-ministry, finding that around half combine high potential with manageable complexity.

Structural reforms have stabilized the situation today, but the path to long-term growth remains uncertain and fragile.

Within this context of broader turmoil, the current government has identified the digital economy as a key component of economic recovery – what Dasun Hegoda, Sri Lanka’s Technical Advisor for Digital Public Infrastructure (DPI), called “an opportunity to reimagine where we want to be in five years”.

Dasun Hegoda, Sri Lanka’s Technical Advisor for DPI, presenting at the GovMesh 4.0

In particular, the National Digital Economy Strategy has earmarked DPI as the foundational layer for realizing Sri Lanka’s vision of a robust, export-oriented digital economy that can sustain economic growth, create new employment opportunities, and deliver trusted public services.

At GovMesh 4.0, Hegoda shared more about how Sri Lanka is tapping on DPI - defined by the World Bank as “foundational, digital building blocks designed for the public benefit” - to meet its digital economy goals.

From silos to horizontals: how DPI powers the digital economy

The South Asian country has set an ambitious target: for the digital economy to comprise 12 per cent of its total economy by 2030, from 3-4 per cent (US$15 billion) currently.

One third of that would come from digital exports.

This reflects the government’s confidence in its strong domestic IT industry, which saw nearly 30 percent growth from 2023 to 2024.

“The whole idea is not only to be digitally ready within the country, but also to compete regionally within Asia and beyond as well,” said Hegoda.

This is where the DPI approach comes in.

A traditional approach to the digital economy would entail each sector developing its own digital platforms, such as digital identity, digital payments, and data exchange systems – a repetitive and costly approach.

In contrast, introducing secure and interoperable foundational systems can enable use within both government and industry.

This horizontal approach to digital infrastructure means that individual sectors, such as healthcare, need not re-invent the wheel or develop their own methods for interacting with users.

Businesses tapping on these common systems can onboard new customers more quickly and securely, while benefitting from lower costs and compliance burdens.

The reusability of DPI across sectors makes it a cost-effective method for achieving whole-of-society digital transformation, which is particularly helpful when working within financial constraints, noted Hegoda.

By coordinating the development of nationwide DPI, the Sri Lankan government aims to support the private sector in building meaningful products and services, contributing in turn to the country’s digital economy goals.

Adopting open-source solutions

Currently, the Sri Lankan government’s strategy for implementing DPI is through the adoption of open-source digital public goods (DPGs).

This is in part to avoid the vendor lock-in challenges posed by proprietary solutions, said Hegoda.

Sri Lanka is working closely with the Digital Public Goods Alliance (DPGA) to identify the right DPGs to form the foundation of its national DPI.

For instance, the country has adopted the open-source MOSIP for its Sri Lankan Unique Digital Identity (SL-UDI) program, which is due for rollout in late 2026.

To drive adoption of the new digital identity program, the government will be prioritizing use cases which have immediate impact on citizen’s lives and can scale quickly, while ensuring security and privacy, said Hegoda.

“All these DPI solutions can be used by the public sector as well as the private sector. Rather than rebuilding all these capabilities, they can pick and choose these ‘Lego blocks’ to build their applications in these domains,” he explained.

This is akin to the adoption of Singpass in Singapore or Aadhaar in India, where successful adoption was driven by integrating identity modules within both public and private services. These modules include Singapore’s MyInfo, which allows users to pre-fill forms with verified data from government sources.

Likewise, Sri Lanka is also leveraging data exchanges similar to the open-source X-Road to build its National Data Exchange and facilitate seamless data sharing across government agencies.

In the long run, there are plans to ensure the entire architecture is AI-ready and interoperable with cross-border systems, he said.

Setting the right foundations

To ensure the success of Sri Lanka’s DPI rollout, the right foundations have to be in place.

First, this includes setting up the right governing bodies, including GovTech Sri Lanka and the Digital Economy Authority, which will drive coordination across government bodies and oversee the nationwide implementation of digital platforms respectively, highlighted Hegoda.

On the legislative front, the country is finalizing its Cyber Security Bill, which authorities say will be tabled in Parliament in 2026, and its data protection laws are slated to be fully operationalized in 2026 as well.

Other efforts that the country is embarking on across the next five years include the modernization and expansion of the existing Lanka Government Cloud and improving digital literacy and digital adoption across the country.

The government will also introduce a capacity building program to support civil servants and public sector bodies in better governing this infrastructure.

“To get everyone onboard, we are trying to make it incentive-driven for all stakeholders,” he said, pointing to the tangible benefits that Sri Lanka hopes to deliver to citizens, businesses, startups, and government officials through this initiative.

The third time, the answer was the same.

The reason: government contracts don’t pay well enough, and there was better work elsewhere.

Thailand’s Digital Economy Promotion Agency (DEPA)’s Non Arkara presenting at GovMesh 4.0

Not one to take “no” for an answer, Thailand’s Digital Economy Promotion Agency (DEPA)’s Senior Expert in Smart City Promotion, Non Arkara, built it himself.

Speaking at GovMesh 4.0, Arkara admitted that he was not a professional software developer. In fact, he had trained as an architect.

With some working knowledge of Python, JavaScript and using a couple of free ecosystems from Google, he built a suite of digital tools that now serving thousands of users across Thailand and Southeast Asia.

His journey pointed to a structural problem that emerging governments around the world must reckon with: when tech talent flows out of the public sector and procurement rules continue to slow things down, what gets built?

Building the tool the market won’t

Arkara’s agency needed a system to allow citizens to report infrastructure problems, but no vendor would take it on at the price or pace that worked.

This got him to look at everyday tools citizens already use, such as messenger apps like LINE, WhatsApp, and Telegram, as well as building a unified backend system that automatically extracts key information to route real problems and filter trolls.

His effort has paid off. While previously Arkara’s team used to get around 10 queries a day, they now receive around 2,000 per day through this system.

For the civil servants, he has created a dashboard that lets them categorize the complaints, assign them to the right teams and send status updates to citizens, all in one click.

This turned a simple reporting tool into a full feedback loop.

Laying artificial intelligence (AI) on the system has also allowed civil servants to analyze the resolution times, workloads, and overall team performance, later then tying these metrics to staff incentives.

Arkara has open-sourced the whole stack, instead of monetizing it.

“I decided to open-source it completely... it’s better to give it away for free.”

“We live in a world where if we just keep it to ourselves, patent everything proprietary, we’re not going to get different government,” he said, effectively turning this from a one-off solution into a replicable public infrastructure.

In another interview with GovInsider, Arkara also discussed his role in introducing the Innovation-as-a-Service model to local governments in Thailand and ASEAN, as a means to tackle resource constraints and rigid, pre-made solutions offered by the market.

A model to follow, or simply a workaround?

The deeper question hidden behind Arkara’s initiative was whether a bottom-up movement, powered by low/no-code tools and AI, represented a model to be encouraged across public service, or should it be seen as a workaround that masks a problem that governments should be fixing at the root.

What does it take for governments to build sustainable digital infrastructure?

On one hand, what Arkara has done, at very low cost, demonstrates that Govtech implementation is possible even in environments where formal digital capacity is thin.

On the other hand, the tools built by one talented and motivated person, or even a team, may not survive his departure or make a digital government strategy.

While his approach worked, it also underscored a tension that Arkara does not shy away from.

“I’m operating as a one-person service for now,” he said, adding that cybersecurity hardening, formal authentication and team oversight are under his purview.

Responding to an audience’s question about data quality and the reliability of underlying data feeding the dashboards and tools, Arkara emphasized the importance of human review and continuous usage to determine the data quality.

What Thailand’s experience makes clear is the gap between what governments need and what the market will supply.

How public agencies respond to the gap, whether by revamping procurement rules, building in-house tech capabilities, or simply relying on the individual’s initiatives, will shape the quality of public services for years to come.

This week, there are new digital bodies springing up in governments across the world, including ongoing work in Japan to set up a new AI reform team to review rules that could slow the testing and deployment of emerging technologies (the UK AI Growth Lab is poised to do something similar). We also look at the progress on the German Federal Modernization Agenda, and the UAE’s dedicated Artificial Intelligence and Data Authority set up to boost its bid to “build the government of the future”.

Our main stories this week:

• OECD’s Digital Government Outlook 2026

• Indonesia’s largest AI experiment barely has a safety net

• Germany plans to make public administration fit for the digital age

• From digital government to digital transformation

As a reminder, we have recently begun to expand our digital government coverage away from just Substack and onto Linkedin, Twitter and Instagram. For original content, audience engagement, and roundups of all our pieces, feel free to follow us over on those channels. Also check out our sister newsletter, interweave.africa.

Digital Government Outlook 2026 - OECD

Recognizing the increasing expectations from citizens that governments act quickly, adapt to change and respond effectively, the OECD have released a global Digital Government Outlook - presenting results from their multiple digital indices.

• 92% of OECD countries now have an AI-in-government strategy, but this is not always applied critically. 75% of countries do not evaluate whether investments in digital technologies and AI deliver the intended results.

• There is a widening gap between rising expectations and institutional pace - citizens and businesses expect speed, adaptability and responsiveness, which traditional institutional bodies are struggling to keep up with.

• The report identifies five priority shifts needed to move from ambition to performance, including scaling uptake of digital public infrastructure (DPI) and strengthening data governance for coherence and reuse.

Indonesia’s largest AI experiment barely has a safety net - Hanif Abdul Halim, Asian News Network

Since 1st June, Indonesia's GovTech platform has integrated data from eight ministries, covering 270 million citizens with a single AI-powered system. But it has done so without any binding AI risk taxonomy, mandatory pre-deployment safety standards or independent audit mechanisms.

• Indonesia’s National Economic Council has confirmed that the GovTech platform has achieved 80% connectivity across eight ministries and agencies, integrating their data into a single AI system for the first time in the country’s history.

• However, a lack of governance around the integration has created concerns relating to accountability and legal responsibilities in the event of something going wrong.

• Such a challenge is far from a mere hypothetical. A previous ransomware attack in June 2024 on the country’s Temporary National Data Center compromised data across more than 200 government institutions for days, and no single ministry or agency was ever held legally responsible. This new architecture will be 10 times the size.

Germany plans to make public administration fit for the digital age - Newsworm

Germany's digital transformation has gathered pace over the past couple of years, largely thanks to its new(ish) digital ministry. A couple of weeks ago, its Federal and State governments reached a new agreement in Berlin on a comprehensive state modernization program.

• The Federal Modernization Agenda, with more than 200 measures, sets out to digitize correspondence, ID and passport applications, address changes, and vehicle registrations by default.

• A “Germany app” will consolidate services in one place, with a voluntary digital wallet holding a “digital twin” of the ID card from 2027, later expanding to birth certificates and driving licenses.

• There are further planned reforms around the justice system - with permits and payments set to become more digital.

From digital government to digital transformation - Manuel Aguilera, IIPP

Aguilera writes up an IIPP Forum conversation between David Eaves and Emily Middleton (Interim Director General for Digital Transformation in the UK Government), which looks at the role of the UK’s Government Digital Service today.

• Digital government is the thinnest, newest layer of the state. Eaves’s framing for the conversation is that digital teams often overstate their centrality, with older layers of the state resisting change more than the digital reformers assume.

• Within the UK government, GDS plays three distinct roles as architect (setting standards), builder (of platforms like the UK’s One Login) and reformer (of culture and roles). Progress will depend on translating that digital work into the “older language” of the state, especially money.

• The hardest barrier, says Middleton, is cultural rather than technical. Middleton refers to it as “dark matter”: procurement rules, pay frameworks and career paths.

Our Take: This is the clearest articulation yet of what "GDS 2.0" is actually trying to be. This framing matters beyond the UK as it's a template that other digital ministries (Germany's BMDS included) will probably borrow from.

GovTech News in Brief

Estonia intends to recognize AI agents with digital IDs - Thomas Claburn, The Register

Estonian PM Kristen Michal has backed a proposal from the Eesti.ai advisory board to issue "AI ID codes", state-recognized digital identities for AI agents acting on behalf of people, companies or institutions. The codes would define scoped, auditable permissions under which agents no longer need to borrow their owner's full credentials to act.

The West can learn from Ukraine’s AI vision beyond the battlefield - Andrew D’Anieri, Atlantic Council

Ukraine is bringing its AI ambitions into public services, with Diia.AI showing how governments can make services more proactive, accessible, and responsive.

Tasmania removes 468,000 driver’s license photos from national facial recognition database - Pulse Tasmania

Tasmania has recalled hundreds of thousands of driver’s license photos from a national facial recognition system as it reviews the privacy and legal basis for sharing biometric data.

Japan to launch AI regulatory reform - Oh Kuemin, The Asia Business Daily

Japan is setting up a new AI and digital reform team to review rules that could slow the testing and deployment of emerging technologies.

Taiwan’s first digital minister on building a trustworthy public system in the digital era - Sol Gonzalez, GovInsider

Taiwan’s Audrey Tang argues that digital government should be designed around civic participation, social trust, and transparent systems rather than top-down tools.

UAE creates dedicated artificial intelligence and data authority to “build government of the future” - Jack Aldane, Global Government Forum

The UAE is set to establish an Artificial Intelligence and Data authority, consolidating its current AI, data and digital transformation portfolio across departments.

AI tool to slash planning decision times as government accelerates push to build 1.5 million homes - The Rt Hon Ian Murray MP and Matthew Pennycook MP, Gov.UK

The UK government is currently trialing AI-native solutions that will halve the time spent on planning applications. Extract, an AI solution that digitizes planning documents, providing planning officers with relevant data points, is currently being rolled out across the country’s local authorities.

The Theory Behind the Practice

From AI assistants to AI agents: a new operational layer for digital government - Yanez Rodriguez, Fabricio Lenin, IADB

Agentic AI presents a step-change for the public sector. While early movers like Estonia have already started trialing Agentic AI for effective data sharing, other governments need to rethink & evaluate their existing processes and infrastructure to move beyond simplistic adoption.

Public sector AI is advancing faster than public sector data - Will Lowe, The Observer

Governments need to prioritize the development of high-quality, accessible and integration-ready data infrastructure for building the next generation of digital public services, argues Will Lowe.

From hundreds of miles above the Earth, satellites can see so many things – from warming oceans to shifting city skylines. But this data is silent without someone to interpret it.

To make it speak, the Japanese Aerospace Exploration Agency (JAXA) is going back to the basics, turning to the oldest tool in human history: storytelling.

“It’s not the satellite data that people want, but the information that people can extract from the data,” says JAXA’s Advisor to the Director, Dr Naoko Sugita, at GovMesh in Singapore.

She shares how the Earth Observing dashboard has evolved from a simple viewing tool into a shared playground, empowering non-space stakeholders to use satellite data to study and solve the world’s most pressing problems.

Making satellite data relevant for the rest of us

The dashboard was first created in 2020 to monitor the worldwide impacts of the Covid-19 pandemic using earth observation data from ESA, NASA, and JAXA.

Together the space agencies worked then - and still do now - with the common goal of making earth observation data more accessible for the public.

“If I can’t use it, other people outside the space agencies can’t use it too,” she says, reflecting on her role as a policy person at JAXA rather than a scientist or engineer.

JAXA’s Advisor to the Director, Earth Observation Research Centre, Dr Naoko Sugita, presenting at GovMesh 4.0

The dashboard was built within three months and delivered on a free, web-based platform, enabling access to more than 850 open datasets across the three agencies, she tells the participants.

In 2022, it expanded its data scope to cover more areas related to environmental change.

In addition to Covid-19 data, the new areas were in atmosphere, agriculture, cryosphere (Earth’s ice), oceans, biomass, economy, and extreme events.

As of 2025, the dashboard had been used across 130 countries, reflecting expansion across Africa, small states, and emerging regions.

Static archive to living tool

While the dashboard was originally intended only to facilitate earth observation data use among citizens, the dashboard has today become a tool aimed at working towards enabling open science.

What this means is encouraging participation from domain experts and other stakeholders to contribute their own narratives and evaluate the values displayed.

“Initially we created the stories within the three agencies, but we’ve been opening this up to like researchers and graduate students. So, they’re creating these stories and we verify it,” she explains.

To make it easy for others to build stories that matter to them, the dashboard uses indicators in accessible language - like CO2 daily, soil moisture, and lake quality - rather than technical metrics related to satellite data.

“Even if you don’t have any specific knowledge about handling satellite data, you can still use it and explore narratives on societally relevant topics or create new ones,” she explained.

Another key element of open science was positioning the dashboard as a lab for others to plug in.

In 2022, JAXA launched the “JAXA Earth API” to make it easier for developers and researchers to use JAXA’s satellite data easily in programming environments like Python and JavaScript.

To the same end, the team recently organized a hackathon with Cyber University, where students used the “JAXA Earth APIs” to build innovative solutions in just two and a half days.

“This proves the strength of the application programming interfaces (APIs) and dashboard to make satellite data accessible. We would like these data to be used more”.

The dashboard is also integrated with Jupyter Notebooks, which allows researchers to run analyses themselves in order to verify the results or tweak the parameters of their own studies.

AI for greater accessibility

Dr Sugita sees artificial intelligence (AI) as a bridge between raw data and practical solutions.

Firstly, analytical AI could be used to detect anomalies and categorize satellite imagery. For example, turning ambiguous pixels into clear categories like “forest” or “urbanized area.”

Secondly, generative AI (GenAI) could be used as a consultant for the user. JAXA is looking to incorporate a GenAI layer on their own dashboard to allow users to ask text-based questions.

“The ideal way is to be able to ask the dashboard a question like ‘When should I plant my rice to get the best harvest?’,” she explained.

By combining past insights with real-time satellite data, the AI could provide actionable responses for the user.

When Dr Sugita concludes her presentation, several participants share how their own governments have already integrated open-access earth observation data to monitor shifting terrains and inform national planning.

The takeaway is that for data to be a truly transformative tool, it requires a human-centric infrastructure.

Through an accessible dashboard, APIs, AI and community engagement, governments can turn technical metrics into meaningful tools for other users to tackle broader societal challenges.

Manuel Kilian presents the Agentic State concept at GovMesh 4.0

AI agents provide governments, he told the leaders convened for GovMesh 4.0 in Singapore, with “the brains and the hands”.

The intellectual forces that are large language models are equipped with tool use, memory, planning and orchestration – able to follow workflows and function autonomously.

“The brain alone is not so useful, and the hands alone are not so useful”, Kilian told participants.

“But through that combination, AI agents are able to receive reason and act with minimal supervision to get an outcome”.

In an interview alongside GovMesh with Civic Punks’ Derek Alton, he describes agents as digital co-workers that will help civil servants by taking over part of their work, without taking it away.

The 12-component architecture of the Agentic State he presents is daunting in its breadth at first glance, but he reassures participants that much of what is needed to achieve it is inherent within governments.

“Government is a bundle of workflows”, he says, “regulations, laws, rules… a lot is codified workflows”. Achieving the Agentic state, he argues, means making those workflows automatic and proactive rather than slow, reactive and episodic.

It is a vision that governments are beginning to buy into, most notably in Ukraine, where former Minister of Digital Transformation Mykhailo Fedorov explicitly embraced the vision of moving towards the agentic state.

The data trends for an Agentic State

Within the contexts of data-driven government, the principal theme of GovMesh, Kilian identified several trends enabling the Agentic State.

As he does so, he anchors on a concrete use case – the birth of a child.

The architecture for the agentic state. Image: GGTC Berlin

The last thing a parent needs when dealing with a newborn is government bureaucracy.

But there are numerous necessary points of interaction with the state needed after childbirth – registering the birth, or signing up for childcare benefits, for example.

In an Agentic State, the idea is that an agent – with the knowledge of the individual context of family – can trigger these necessary services automatically.

To do so, said Kilian, they need to adhere to three principles:

• Data as self-describing assets with rich metadata: When a birth is registered, or any other data point gathered, it comes with a semantic and metadata context. Is a birth a first or second child, or is this birth verified by a hospital, for example? In shifting to thinking about a data point as an asset for AI agents, not just a row in a database, Kilian encouraged the other participants to deliberately cultivate metadata context, provenance and permissions.
  

• Ecosystem-wide governance and stewardship: “A life event can trigger a cascade of public services”, Kilian said. A birth registry, for example, might need to be cross-checked against family income to test their entitlement to child databases. Governments are trying to find a balance between open-by-default ecosystems, where data and metadata can enable innovation and scrutiny, and a privacy-first mindset that guards data access.
  

• Privacy-preserving technologies for sensitive data: As governments attempt to find this balance, Kilian encouraged a privacy-first mindset. “To see if a family is eligible for housing support”, he says, “an agent doesn’t have to access the entire income situation of a family. It just needs to know whether it is above or below a threshold”. Agents need to have access to minimum viable data, able to expose insights without exposing individual records.

Taken together, said Kilian, these principles lay the foundation for a vision of “agents as data scientists”.

Practically this means agents automating data-ops end-to-end. This makes agents able – for example – to run policy simulations and explore what might happen if a government changes a threshold to child benefits.

A second white paper, written by the GGTC together with the World Economic Forum (WEF), aims to decode all of governments’ activities into 70 discrete workflows (see here).

It is a complicated balancing act, but a bold vision of what the next generation of public services might look like.

This week, our focus remains on these dynamics. We look at NATO’s procurement of digital software, as well as Poland and The Netherlands’ respective sovereignty responses when it comes to domestic digital infrastructure. Elsewhere, the UK Home Affairs Committee has delivered a scathing verdict on the Digital ID “BritCard” rollout, a warning lesson to any other governments working on similar products.

Our main stories this week:

• NATO’s digital back end could fall apart without change - David Eaves, Foreign Policy

• Mandatory digital ID launch called a ‘fiasco’ by scrutiny committee - Jennifer McKiernan

• Making Agentic AI work for government: a readiness framework - World Economic Forum

• Netherlands blocks US takeover of vital digital supplier - Pieter Haeck, Politico

As a reminder, we have recently begun to expand our digital government coverage away from just Substack and onto Linkedin, Twitter and Instagram. For original content, audience engagement, and roundups of all our pieces, feel free to follow us over on those channels. Also check out our sister newsletter, interweave.africa.

NATO’s digital back end could fall apart without change - David Eaves, Foreign Policy

Writing in the aftermath of February’s Operation Epic Fury, Eaves argues that cloud capabilities have become a core part of allied warfighting, rather than a back-office IT concern.

• During the US campaign in Iran, peak daily use of one Pentagon “AI targeting system” reportedly jumped a staggering 4,400%, exposing how dependent live operations are on rented compute and other software.

• Eaves points to the familiar gap between NATO’s stated digital ambitions and reality - a proposed alliance-wide cloud and data-sharing “digital backbone” by 2030 contrasted with a slow, fragmented pace of action.

• This piece lands at a time when NATO is leaning further into commercial providers, sharpening questions about resilience and dependency.

Our Take: This is the defense-flavored cousin of the sovereignty debates we keep returning to, especially in Europe. The uncomfortable thread here is that “wartime tempo” and “careful procurement” pull in opposite directions. Eaves is right that compute is now a weapons system in itself and not simply a piece of the puzzle, and procurement is yet to catch up.

Mandatory digital ID launch called a “fiasco” by scrutiny committee - Jennifer McKiernan, BBC News

The UK’s Home Affairs Committee has delivered a blunt verdict on the government’s handling of “Britcard”, concluding the original mandatory plan was rushed and poorly explained.

• Committee chair Dame Karen Bradley called the early rollout “nothing short of a fiasco”, noting the announcement “came out of the blue” with no real policy groundwork behind it.

• The compulsory element was quietly dropped in January after a public backlash that included a petition with nearly three million signatures.

• MPs also flagged that mandatory digital right-to-work checks remain on the table, effectively forcing those without a passport into the digital ID system anyway.

Our Take: This is a textbook case of how clumsy sequencing can sink a policy. The committee’s “staircase, not a slippery slope” line is a neat one, and other governments eyeing their own ID rollouts would do well to take note of the critiques being aimed at the UK government before announcing similar policies.

Making Agentic AI work for government: a readiness framework - World Economic Forum

The WEF’s second GovTech report with Capgemini and GGTC Berlin sets out a department-agnostic way of working out where agentic AI is actually ready to do useful work. It scores readiness across 70 government functions, rather than taking a ministry-by-ministry approach.

• The headline framing is a potential $9.8T public-sector opportunity by 2034, with roughly 90% of institutions reportedly planning to explore or deploy agentic AI within the next two to three years.

• More usefully, the analysis finds that around half of the functions WEF assessed combine high potential with manageable complexity, meaning there are plenty of places for governments to begin.

• The emphasis throughout is on human oversight, strategic sequencing and avoiding the all too familiar trap of ambitious pilot phases which never scale.

Netherlands blocks US takeover of vital digital supplier - Pieter Haeck, Politico

After months of public debate and protest, the Dutch government has blocked US firm Kyndryl’s acquisition of Solvinity, the cloud provider that hosts DigiD (the country’s digital identity backbone).

• State Secretary for Digital Economy Willemijn Aerdts told parliament that the national investment-screening authority had advised blocking the deal as a “possible risk to the public interest.”

• The core fear was that DigiD data could fall under US jurisdiction and be demanded by American authorities, a common sovereignty concern given fresh urgency by the current transatlantic mood.

• The decision followed parliamentary motions, citizen legal action and sustained campaigning from groups like Privacy First, with the Solvinity contract having earlier been extended to avoid a risky rushed switch.

GovTech News in Brief

Luxembourg opens tender for AI-generated content detection tool - Masha Borak, Biometric Update

Luxembourg is seeking vendors for a deepfake detection platform aimed at helping its media regulator monitor AI-generated and manipulated content.

Why Italy’s DPI doesn’t feel like infrastructure - Ottavia Porcellini, GovInsider

Italy’s digital public infrastructure has expanded through national platforms such as SPID, pagoPA, and IO, but still faces challenges around consistency, adoption, and user experience.

Nect selected to provide digital ID alternative to BundID in Schleswig-Holstein - Masha Borak, Biometric Update

Schleswig-Holstein will offer Nect Wallet as a digital ID alternative for accessing public services in an attempt to boost participation in a country where digital ID adoption lags.

Presentation of “Albanian Digital Solutions”, a new step in the digital transformation of public services - Government of Albania

With Albanian Digital Solutions, Albania is looking to build stronger in-house capacity to modernize public services and move the administration closer to the AI age.

The Philippines made digital inclusion land by making communities co-own it - Si Ying Thian, GovInsider

In rural communities in The Philippines, the Smart Villages and Smart Islands initiative is driving innovation and resilience through collaborative digital transformation.

Poland will introduce “sovereignty test” for government tech purchases as Tusk warns AI dependency has reached dangerous proportions - Alina Maria Stan, NextWeb

Poland has become the latest European country to add sovereignty rules to its procurement practices, including mandating annual IT independence reports to assess the state of sovereignty.

Finland is Europe’s most digitalized country - but older people are still left behind - Jeff Hearn et al, The Conversation

With Finland’s ombudsman for older people finding that 40-45% of those aged 65+ lack digital skills, Hearn explores the demands of digital inclusion in the most digitized country in Europe.

How digital public services are changing daily life in Central Asia - Tamila Olzhbaekova, The Times of Central Asia

Kazakhstan, Uzbekistan, and Kyrgyzstan are moving core public services into mobile apps and online portals, producing as many unanswered questions as concrete results.

The Theory Behind the Practice

Digital sovereignty is in the fine print - Gabriela Ramos and Emilija Stojmenova Duh, Project Syndicate

Ramos and Stojmenova Duh suggest that beyond megaprojects and eye-catching announcements, governments can build leverage on digital contracts through public procurement and services rules.

Digital embassies and AI sovereignty: building resilient states beyond borders - Tony Blair Institute for Global Change

As AI infrastructure increasingly becomes a question of national security, the Tony Blair Institute explores whether and how digital embassies might provide safeguards against domestic disruptions.

Last year, interweave told part in Civic Punks’ new podcast, covering the latest in digital government innovation around the globe, hosted by Derek Alton.

In this format, panelists from different regions – Africa, Asia, Europe, North America and Latin America – each bring a digital government story from their region to the conversation, picking it apart and outlining the key trends.

This time, we were talking about sovereignty. Other participants included:

• Angela, who talked about LatamGPT – the first LLM trained in Spanish using data from Latin America and the Caribbean, raising questions about regional collaboration for sovereignty, and who funds ongoing adaptations and upgrades to so-called “sovereign models”

• Brian, who spoke about ongoing work in the US to build trust in AI among civil servants, and about research into how AI rollout is uncovering demand in some areas by reducing bottlenecks in others. Choosing to implement AI therefore becomes not a question of outright efficiency, but of recognizing where it creates and reduces demand.

When is Sovereignty not Sovereign?

Luke from team interweave brought a couple of stories to the podcast. First was some news from France on how the government is set to ditch Microsoft Teams for a homegrown equivalent, Visio, a move that follows similar experiments from regional administrations in Germany and the Netherlands.

But elsewhere, the story of the Netherlands and Solvinity offers a fascinating edge case in what it means to be sovereign. The case revolves around Solvinity – a company building digital services for public sector actors such as the City of Amsterdam and the Dutch Ministry of Justice, who in turn aim to procure from Dutch-headquartered companies (on sovereignty grounds).

Since beginning work with the Dutch public sector, Solvinity has been subject to an attempted acquisition by American company Kyndryl, the world’s largest independent IT infrastructure services provider (and a spinoff from IBM). The Dutch government found out only a couple of days before the decision was made that this would be happening, prompting huge panic in digital circles.

Kyndryl has attempted to acquire Solvinity, responsible for making Dutch digital services

In the last couple of days, after the release of the podcast, the Dutch government have blocked the acquisition over public safety concerns. If it had been allowed to go ahead, the acquisition could have left Dutch population data subject to the US Cloud Act and so-called kill switches – the very problem that the authorities had tried to avoid in their procurement choices.

This is not just a Dutch – or American – problem, with a parallel event taking place in the UK, where a British company working on the National Health Service app was acquired by a Canadian one. Both these instances offer an interesting case of the limitations of sovereignty, which will no doubt shape on of the key technology debates of our time.

A “1 million AI prompters program”, run in partnership with the UAE government, is looking to train one million people in artificial intelligence (AI) prompting skills.

The country has also set the target of US$1.5 billion in AI software products and services by 2030, with 10 per cent of e-services based on AI.

Technology has become a key enabler of effective public service delivery and sits at the heart of the work led by Aziza Umarova and her team at the Agency for Strategic Development and Reforms under the President of the Republic of Uzbekistan.

The team initially developed an open-source geoportal to track the delivery of critical cross-government priorities in areas such as water management, education, and healthcare.

They help policymakers visualize disparities, monitor progress, and enhance data-driven policy making across institutions.

Through the platform, Umarova told participants gathered for GovMesh 4.0, the government can monitor the development of priorities across the country and improve public investment management – by connecting strategic metrics to budgets in one holistic view.

“It is a major shift across government towards data-driven decision-making”, she said.

Building the fundamentals for data-driven government

“For years, a lack of reliable data had been a major barrier in Uzbekistan” Umarova told GovInsider during an interview in December, “[with the geospatial platform], we decided to turn the table”.

The result is an open-source platform developed around location intelligence. Its design is supported by good legislation and citizen collaboration.

A new draft data law clarifies ownership and stewardship, mandates inter-agency sharing, protects privacy without blocking use, and enables official statistics and admin integration.

“We needed a new legal framework to prevent excuses for ministries not to share”, she shared.

Inspired by a Dubai law that says data belongs to a state, not a minister, this legislation mandates sharing of data across government.

“People cannot just not share because they are doing badly,” she said.

The draft law also introduces a data council reporting directly to the President, with data stewards at the Deputy Minister level in every single ministry, similar to Ukraine’s CDTOs.

Students participate in “mapathons” for Uzbekistan’s national geoportal. Image: Umarova’s presentation

But perhaps more important, and impressive, has been the mobilization of young people to support the platform’s development, a so-called “mapathon”.

It is visible to citizens through mapping, generating a collective buy-in which enables a bottom-up gathering and validation of data.

“This was first tested with water-access mapping”, Umarova said, “where 120 students produced a full household-level dataset for an 8,000-household district”.

Through the students’ mapping, the government was able to show areas of lower water access and poorer water quality, all without any formal infrastructure.

These young people are a valuable resource.

“In a country of 38 million people, 60 per cent are below 30”, she reminded the participants.

“With a slogan of ‘my country, my map’, we have been able to work with university students to collect data on schools and public institutions”.

This “participatory sensing” collects micro-data that is then fed back into national data systems.

Not only does this improve data quality, but it builds a data consciousness throughout the population.

Turning data into policy results

Good data is a tool for good policy, but not an end point on its own.

Uzbekistan has also been working to shift the way the government thinks and operates. “We have begun to use evidence instead of anecdotes”, she said.

Ongoing use cases for the geospatial platform. Image: Umarova’s presentation

As she told GovInsider previously, “success for us means when a policymaker stops saying ‘I think’ and starts saying ‘the data shows’”.

Collecting data on social facilities has allowed the government to better determine approaches for planning based on changes in demographic data.

Knowing how many children are approaching the age of educational enrolment, or migration patterns in a given region, allows the government to make better calls on where to open a school or a sports club.

Elsewhere, Land Use Land Cover demonstrates changes in pastures and forestry on a map are demonstrating the effects of climate change over time.

A further byproduct of the geospatial platform has been creating a cross-government view by gathering indicators around health, development and education all in one place.

“Ministers are required to place themselves in each other’s shoes”, Umarova said, “it doesn’t matter if your portfolio is construction, education or healthcare. You see everything, not only yourself, on the platform”.

It is a model of government that requires a whole-of-government approach and ability to mandate government bodies to share data.

But with data formats like satellites becoming increasingly accessible (as Japan’s Dr Naoko Sugita shows elsewhere in GovMesh) – it is a model that holds potential for many others to learn from.

“Everything is possible”, Umarova closed with, “but the oil for this, the fuel for this, is a good quality of disaggregated, structured and cleaned data sets”.

By prioritizing open-source technologies, strengthening local capacity in data analytics, and embedding citizen-centric design across government systems, Uzbekistan’s reforms aim to place people - rather than bureaucracy - at the center of governance.

So too in Europe, where much of the public debate around the European Union’s AI Act has centered on these foundational model risks and capabilities. Italy is beginning to move in a different direction.

Italy’s Agency for Digital Italy (AGID) is advancing a broader, full-stack approach to AI governance, one that looks beyond models to the infrastructure and systems that support them.

Its framework spans three complementary documents covering the adoption, development, and procurement of AI systems in public administration, with the development and procurement guidelines having closed public consultation in April 2026.

To better understand these guidelines, interweave’s team spoke with Mario Nobile, Director General of AGID, earlier this spring.

A full-stack approach to AI Governance

AGID’s framework argues that governments should govern every layer of the AI stack: energy, chips, infrastructure, models, and applications.

Rather than evaluating AI systems solely on model performance or risk classification, the framework encourages assessment across broader socio-economic and operational dimensions, including energy use, infrastructure resilience, interoperability, scalability, and long-term institutional capability.

AGID’s approach to governance looks across the AI stack, not just at model safety. Source

This stack-wide view extends to how AI systems behave once deployed.

The framework introduces an autonomy classification ranging from L0 to L5, modeled loosely on self-driving car taxonomies, where agentic systems follow an observe-decide-act loop and higher autonomy levels trigger stronger governance and traceability requirements.

Italy places most current deployments between L2 and L3, with L5 reserved for research; a way of tying the abstract question of “is there a person behind this?” to concrete oversight obligations.

As a result, this integrated approach to AI means asking governance questions that are often overlooked in mainstream AI debates:

• What happens if preferred compute infrastructure becomes unavailable?

• Can systems operate in lower-capacity environments as fallback options?

• Can governments migrate systems without excessive technical or financial costs?

These are not simply technical procurement questions. They are questions of resilience, state capacity, and technological sovereignty.

As Nobile explains, the goal is to avoid lock-in “not only from the point of view of the model, but the entire system.”

This approach also reshapes procurement.

By emphasizing portability, reversibility, interoperability, and lifecycle sustainability, the framework seeks to ensure that public administrations retain meaningful operational flexibility even after adoption. In practice, this reframes AI governance as a question of institutional agency and strategic autonomy, not merely model safety.

Multiplicity over monopoly

At the center of the framework is what could be described as a “multiplicity principle”: governments should preserve the ability to operate across multiple AI architectures, providers, infrastructures, and deployment environments simultaneously.

Rather than treating open-source and proprietary AI systems as opposing ideological camps, the guidelines assume governments will work across multiple architectures at once: proprietary, open-weight, cloud-based, on-premise, and hybrid systems.

Italy’s AI-orchestrating infrastructure aims to support model multiplicity. Source.

“The future will be multiplicity,” Nobile tells interweave.

This is a subtle but important shift. Much of today’s AI governance debate implicitly assumes convergence around a small number of dominant providers or architectures. AGID’s framework instead treats dependency itself as a governance risk.

Under this approach, models are evaluated less on ideological preference and more on functional criteria such as governability, explainability, portability, and long-term operational performance. Italy may be among the first governments attempting to operationalize AI pluralism as state policy.

The framework also introduces the idea of a “Levelized Cost of AI” (LCOAI): evaluating AI systems not only by upfront performance, but by their full operational lifecycle, including infrastructure spending, retraining, governance overhead, migration expenses, cloud consumption, and transition costs.

In procurement terms, this shifts attention away from short-term capability gains toward long-term dependency risk and institutional flexibility.

Governance, trust, and institutional capability

At a broader level, Italy’s framework also attempts to reduce public anxiety around AI adoption by strengthening institutional capability rather than slowing deployment altogether.

“I want everyone to feel better, with confidence that no one is replacing them,” Nobile says.

The guidelines explicitly frame AI systems as tools intended to support workers rather than replace them. They aim to build institutional awareness around how AI may affect workflows while ensuring that human oversight and administrative understanding remain central.

At a time when much of the AI governance debate remains focused on model risks alone, AGID’s framework offers a different vision, one centered on institutional agency, operational flexibility, and technological sovereignty.

Moving away from asking governments to simply trust the systems they adopt, this Italian framework emphasizes their ability to continue understanding, governing, and adapting those systems over time.

Building institutional capacity to classify operator roles, calculate lifecycle costs, and enforce portability clauses requires skills and bandwidth that most public administrations are still developing. Italy’s guidelines acknowledge this but the value of a serious framework is in what it makes possible to ask for, to measure, and to expect over time.

That, quietly, is what changing the terms of the conversation looks like.

This week, our focus is back on how governments are responding to and adopting AI, reflecting on the UAE’s planned shift to the Agentic State and Singapore’s Mythos-induced cyberresilience meeting. Elsewhere, there is lots happening in the digital ID world, from Indian ID credentials to a Saudi rollout of digital ID for Hajj pilgrims.

Our main stories this week:

• Indonesia moves towards the right to be forgotten

• MAS meets with financial leaders to expand cyber resilience in the face of AI’s advancements

• UAE plans to run 50% of government on Agentic AI within two years

As a reminder, we have recently begun to expand our digital government coverage away from just Substack and onto Linkedin, Twitter and Instagram. For original content, audience engagement, and roundups of all our pieces, feel free to follow us over on those channels. Also check out our sister newsletter, interweave.africa.

Indonesia moves to strengthen right to be forgotten in Human Rights Law revision - OpenGovAsia

Indonesia is taking steps to further enhance digital rights for citizens, including the ability to wipe your digital footprint clean, or “the right to be forgotten”.

• The right to be forgotten allows individuals to request the removal of personal data or documents from online platforms and search engines when the information is inaccurate or outdated.

• Motivation for this legislation is partially motivated by the desire to allow individuals to clear their name if they have been found not guilty of a crime, when outdated information can remain accessible for years after a matter is officially settled.

• Indonesia’s Human Rights Minister noted that a formal process for reviewing such cases would be created, and it would not just occur automatically.

(Singapore) Chiefs of financial institutions met MAS to discuss measures against AI threats following Mythos’ limited release - Sarah Koh, The Straits Times

Following the groundbreaking “release” of Anthropic’s cybersecurity-focused Mythos model, Singapore’s central bank is working with financial institutions in the country to pre-emptively take measures for a potentially more insecure cyber world of tomorrow.

• The CEOs of major financial institutions met with the Monetary Authority of Singapore to discuss collective action on this problem.

• Certain financial institutions in the US, such as JP Morgan, have already gotten access to the model, which was found to discover major vulnerabilities in every major operating system and browser.

• The meeting was followed by Singapore’s cybersecurity commissioner sending letters to the senior leadership of critical information infrastructure owners in the country (including aviation, healthcare, media, and finance), encouraging them to review their own cyber-risk posture.

UAE plans to run 50% of government on Agentic AI within two years - MIT Sloan

The UAE is targeting half of governmental services and operations running on agentic AI within two years - making it the first state in the world to adopt such a model.

• The announcement from the UAE’s President outlined specific performance metrics: speed of adoption, quality of implementation, and the extent to which agencies can redesign workflows around AI capabilities.

• Oversight for the transformation will sit with UAE Vice President Sheikh Mansour bin Zayed, while execution will be coordinated by a taskforce chaired by the Minister of Cabinet Affairs.

• It is hoped that the program will strike at the core government challenge of institutional capacity, with civil servants retrained and positioned as operators and supervisors of AI systems.

GovTech News in Brief

Nepal introduces government e-marketplace, with direct online procurement and digital purchasing system - TechPana

Nepal has unveiled a new e-marketplace system for streamlining public procurement, designed to allow government bodies to purchase goods and services on the platform up to a certain limit.

(Thailand) New push seeks to streamline digital policy - Komsan Tortermvasana, Bangkok Post

Thailand’s digital ministries have released a proposal to help unify the country’s digital direction under a single policy framework, rather than sitting across different administrative bodies as it does today.

Our Take: While we at interweave tend to talk about new initiatives, it’s often helpful to remember that consolidation of services can often be just as - or even more - important in order to promote effective governance.

Kazakhstan government reviews plans to expand AI across sectors under digital strategy - DigWatch

Digital Kazakhstan - the Central Asian country’s initiative for driving long-term growth and economic modernization - is potentially having its mandate expanded with regard to strengthening infrastructure and productivity across all sectors.

Google Wallet supports Aadhaar verifiable credentials in India - Lu-Hai Liang, Biometric Update

India’s Aadhaar system can live entirely within your phone’s wallet from today, allowing for a more secure and privacy-preserving method of digital verification.

(Japan) Government to launch AI pilot program to boost efficiency and encourage tech’s adoption - The Japan Times

Japan’s government is poised to launch a pilot program for its generative AI intelligence platform developed for internal use by civil servants - with an estimated 180,000 workers expected to be eligible.

France Identité app launches sandbox for iOS, proves age check privacy bona fides - Joel McConvey, Biometric Update

France’s ID verification app has now hit its sandbox build for iOS, allowing early testers to experiment with end-to-end flows. The ID will be compatible with the broader European Digital Identity wallet.

Germany's "sovereign AI" champion Aleph Alpha sold to Canada's Cohere in $20B deal - Kai Nicol-Schwarz, CNBC

Cohere is acquiring Aleph Alpha, Germany's flagship "sovereign AI" startup. The deal values the combined entity at $20B and is officially framed as a "transatlantic AI alliance" anchored by Germany and Canada.

The Theory Behind the Practice

Why integration alone cannot create a digital government - Meder Toitonov & Felix Shapiro, GovInsider

This piece from GovInsider describes challenges from Central Asia’s digital governance journey, arguing that simply interlinking systems is often insufficient for improving digital government services. Governments should instead prioritize a “data-first” strategy in small pilots to iron out issues before everything becomes “connected”, the author argues.

Upcoming GovTech Events

ATxSummit / Asia Tech x Singapore 2026 - 20-21 May

Hosted by IMDA (Infocomm Media Development Authority of Singapore), Asia’s flagship GovTech convening will see 4,000+ participants from 50+ countries, with an opening keynote by Josephine Teo, Singapore’s Minister for Digital Development and Information.

Evidence-based policymaking is a priority for governments worldwide. But fragmented data across different agencies remains a persistent challenge in many countries.

Rather than functioning as a shared asset, data can often develop within institutional silos, limiting its strategic value and preventing optimal use.

One of Mongolia’s key efforts to address this challenge has been to pursue data integration through a series of pilot projects that have since evolved into a more systemic approach, as the Director of the country’s National Statistics Office explains during GovMesh.

Ariunbold Shagdar’s team plays a central role in this effort, managing two core registration systems: population and housing, and business registration.

The former is individual-based, while the latter is based on legal entities.

By connecting these two systems, the government has begun to build a more comprehensive picture of the country’s conditions.

“If we connect these systems, we can see a better picture of the country. Every policy and program could be more productive and efficient,” he adds.

From face-to-face to register-based census

One of the most visible impacts of data integration pilots was how it has transformed census operations.

While census collection was previously conducted entirely through face-to-face, the approach has now shifted significantly.

Shagdar shares that by adopting a mixed-method approach for the 2020 Population and Housing Census, the National Statistics Office of Mongolia achieved significant cost efficiencies.

“By utilizing administrative registers for 90 per cent of the data and face-to-face interviews for only 10 per cent, the project budget was reduced from 15.2 billion MNT ($5.4 million) to 4.7 billion MNT ($1.7 million) - a total saving of 10.5 billion MNT ($3.7 million).

“Furthermore, by moving to a fully register-based system, the 2025 census cost less than one billion MNT ($282,064).”

Supported by automation tools such as ETL (extract, transform, load), Mongolia is also exploring the possibility of conducting annual census updates to provide more timely data for policymakers.

Addressing inter-agency resistance

The journey towards integration has not been without challenges, particularly resistance from government agencies and lack of compliance with data standardization.

This is why the Mongolian government has rolled out several policies to mandate data sharing, supported by legal frameworks on statistics, transparency and data protection.

For Shagdar, success depended not only on regulation, but also on leadership and the presence of a trusted institution to drive the initiative.

“There should be leadership. Some organization should lead this initiative,” he notes.

He also highlights the importance of an incremental approach, rather than waiting for perfect conditions.

Instead of waiting for everything to be ready, Mongolia has adopted a “learning by doing” approach by building systems while continuously refining them.

Building the foundations of a data ecosystem

The Mongolian government is also developing a broader data ecosystem, including a national database with multiple layers including aggregated data, customized datasets and data visualization tools.

Metadata standardization has become a priority to ensure interoperability across systems.

“If we want to integrate all government databases, we should standardize metadata,” Shagdar explains.

In addition, Mongolia is establishing data laboratories, expanding access to statistical data, and revising legal frameworks to support greater data openness.

Shagdar emphasized that lasting change depends on demonstrating tangible results.

“We should demonstrate some very good flagship work. Over time, it will work,” he said.

As with previous editions, the event sought to bring together digital government leaders “beyond the usual suspects” that usually convene regularly at international conferences.

This edition also introduced a central theme running throughout the day – data-driven government – and broadened the group of participants from central digital government agencies to include delivery unit leads; national statistics offices and space agencies.

Mongolia’s Ariunbold Shagdar explains the importance of a central interoperable database to the country.

This broader coalition of participants reflected the increasing importance and proliferation of technology across the public sector.

Six governments participated in the event – Indonesia, Sri Lanka, Mongolia, Japan, Uzbekistan and Thailand – together with experts from the Global GovTech Centre in Berlin and UNDP Singapore.

Though the theme was ostensibly data-driven government, the thread that ran throughout the day was moving from data to public value.

From the Agentic State to the rollout of digital public infrastructure (DPI) in Indonesia and Sri Lanka to the use of satellite data, there was a recognition from all the speakers of the importance of telling stories that matter to society and getting beyond technocracy.

As Naoko Sugita, Advisor to the Director of Japan’s Aerospace Exploration Agency, put it, “what people want is not data but information”.

Rolling out data-driven initiatives across government

As is traditional at GovMesh events, the day began with a series of lightning presentations, allowing participants to lead with “their best foot forward” in showcasing a concrete initiative that each of them was working on.

We heard about cross-sector interoperability in Indonesia, the 12 layers of the Agentic State, Mongolia’s interoperable databases and Japan’s use of satellite data to empower policymakers and the public.

Though the focus was on data, participants left each other sticky notes with questions that covered the whole tech stack – thinking about talent, climate change and energy requirements.

This shift in focus towards ecosystem-level thinking mirrors a trend we began to see last November at GovMesh 3.0 in Vilnius, where questions of sovereignty prompted participants to think beyond the service and application layer in their definitions of “digital government”.

Safe and responsible stewardship

Likewise, there was an increasing shift towards thinking about safety and participants’ responsibilities as stewards of data.

The discussions returned throughout the day to questions of accessibility, and data misuse. “What is lost by being fully digital?”, one participant asked after a presentation on automated data-gathering.

Participants chose themes for the afternoon’s Futures exercise.

Continuing the trend of previous GovMesh events, these questions were funneled into an afternoon “futures exercise”, where participants explored digital polarities.

In a quest to imagine “an enjoyable and impactful public service”, the government representatives balanced revolution and incremental change, automation and human judgment, and outsourcing versus building in-house.

Our next GovMesh event in The Hague will take place in June 2026, and continue these discussions with a focus on resilience.

GovMesh 4.0 engaged a visual artist, Wendy Wong, to map out the presentations and the “futures” exercise across the day. You can find the visual maps below:

Individual country’s presentations:

“Futures” exercise mapping the digital polarities:

In a shorter edition than usual this week, the focus is on the EU, and its multifaceted push into all things digital identity. We look at its age verification efforts, cross-border travel, and what this means for sovereignty.

As a reminder, we have recently begun to expand our digital government coverage away from just Substack and onto Linkedin, Twitter and Instagram. For original content, audience engagement, and roundups of all our pieces, feel free to follow us over on those channels. Also check out our sister newsletter, interweave.africa.

GovTech News in Brief

Europe gets serious about age verification online – Laura Carrer, Wired

The EU has begun cracking down on online platforms it deems to be failing to protect minors, having converged upon a “mini-wallet” solution for age verification.

Schengen-wide digital border controls go live – Visa HQ

The Schengen digital border control system has officially gone live, marking the formal end of manual passport stamping for non-EU visitors.

Armenian government pushes to require AI labels on TV content – The Armenian Report

Armenia is the latest country looking to formalize rules around labeling of AI content, this time within the contexts of TV broadcasters.

Europe’s “.eu” hits 3.8 million domains amid push for tech sovereignty, EU Commission – European Commission

The EU’s .eu domain has passed 3.8 million active registrations, with EU officials presenting it as a trusted European digital identity marker and part of the bloc’s wider push for tech and strategic sovereignty.

Indonesia: balancing AI growth with readiness and resilience – OpenGovAsia

Indonesia is positioning itself as a fast-growing AI market, but long-term success will depend on whether the country can match investment momentum with stronger skills, governance, regulation, and inclusive digital development.

Vietnam approves national cultural digital transformation plan until 2030 – Voice of Vietnam

Vietnam has approved a national plan to digitally transform its cultural sector by 2030, with longer-term goals to build a smart digital cultural ecosystem by 2045, expand access, and grow cultural industries.

Pakistan’s government begins providing relief funds through digital wallets – The Express Tribune

Pakistan has begun distributing relief funds through digital wallets, with the government saying that the system will transport subsidies more quickly and transparently to public transport and goods vehicle operators.

Scammers take advantage of Austrian digital ID certificates’ expiry – Masha Borak, Biometric Update

With 300k Austrian digital ID certificates set to expire this year, scammers are pouncing on the opportunity, sending phishing links to gain unauthorized access to users’ devices.

Microsoft pledges USD $1 billion for Thailand AI hub – Mark Tarre, IT Brief

In a meeting with Thailand’s PM, Microsoft announced a $1bn AI hub in the country. This is the latest step in a partnership Thailand has been developing since 2023.

Air New Zealand trials new digital ID – Tara Craig, Airports International

Air New Zealand, the country’s flagship carrier, is the latest airline experimenting with digital ID for streamlining interactions - with the promise of biometric verification and Apple Wallet support for those who submit passport details.

Our Take: While we tend to focus on government agencies at interweave, it’s also important to recognize that tightly-coupled private services like national airlines represent still a huge proportion of how citizens interact with digital services.

UN warns of urgency in shaping responsible AI governance – DigiWatch

UN Secretary General António Guterres has told the UN’s newest scientific panel on AI that its members have a responsibility for shaping AI for the benefit of humanity.

However, a unique challenge exists for the Scottish government when it comes to AI. Whilst it is responsible for delivering health and education outcomes, it does not have the power to regulate the underlying technology used within vital public services. This power sits with the Westminster government in London, and creates a “governance gap.”

The Governance Gap

There is a distinct irony in Scotland’s current predicament. Scotland launched its first AI Strategy in 2021, before the UK’s own national framework was published. But despite this first-mover advantage, Scotland’s voice has been notably absent from high-profile global stages, such as the 2023 UK AI Safety Summit.

Furthermore, a significant strategic divergence is emerging between Edinburgh and London. While the UK Government has favored a sector-led, non-statutory approach to regulation of AI, the Scottish Government’s approach since 2021 has signaled a preference for a more structured, rights-based model.

Although Scotland is not legally bound by the EU AI Act, the government is choosing to align with many of its principles such as classifications of high-risk sectors.

Beyond policy: the battle for social license

The country’s new AI Strategy is an ambitious attempt to fill this governance gap through physical and economic levers. Central to this is an £8.2 billion AI Growth Zone in Lanarkshire, a 9,000-acre site designed to deliver over 3,400 jobs and position Scotland as a European hub for AI infrastructure.

Construction has already begun at the Lanarkshire site with final completion expected in early 2030. Source: The Scottish Sun

Infrastructure projects such as the AI Growth Zone face a critical hurdle in acquiring social licenses. Several large-scale data centre planning proposals have been rejected in Scotland over the past few years, due to concerns about the massive energy and water demands needed to sustain the data infrastructure.

For Scotland to meet the targets outlined in its AI Strategy, it must frame these construction projects as community-first assets, working with local residents and emphasising sustainable data centre designs.

Strategy as pragmatism: control vs. risk

The most telling aspect of Scotland’s new strategy is its focus on an “AI Stack”. By organising action across different layers of AI landscape - such as semiconductor imports, data centres and SME adoption - the Scottish parliament is doubling down on the areas that it has direct agency over.

Because Scotland cannot legislate on AI safety, the government is pivoting its approach to focus on procurement and transparency as its primary tools of digital governance. These are softer instruments than formal regulation but will shape behaviour across the public sector if used consistently.

The most visible expression of this is Scotland’s AI Register, which aims to increase visibility of AI use across all public bodies, moving toward what the strategy calls “transparency floor”.

This register is a publicly accessible database where public sector bodies disclose what AI tools they are using, how they are using them for and what data they rely on. Launched in March 2023, the register is currently voluntary and does little to paint a comprehensive picture of public sector AI use in Scotland.

However, a phased approach is being taken towards mandatory registration, starting with Scottish Government departments.

Alongside this sits a “chain of trust” approach to the wider regulatory environment. By aligning with OECD AI principles and actively advocating for UK regulation that mirrors the EU AI Act, Scotland is attempting to construct a de facto regulatory environment from the outside-in. It cannot write the rules, but it can position itself so that when the rules are written, Scotland’s public sector is already compliant, and its AI industry already has market access.

Taken together, these are not stopgap measures. The success of Scotland’s 2031 vision will not be measured by the number of infrastructure projects they are successful in completing, it will be measure by public trust emerging in the systems being used in hospitals and schools.

By focusing on what it can control, Scotland is quietly building a governance infrastructure around AI that is designed to outlast the legislative gap itself.

Whenever Westminster does act and moves to regulate AI use, Scotland will already be ahead of the curve.

Despite growing awareness of the significance of innovation within government, efforts to drive innovation are often hamstrung by the resistance to change endemic to large, bureaucratic organizations: the bigger the organization, the more challenging it is to effect change.

This can be compounded by other factors, such as the need to prioritize immediate stakeholder needs over nurturing emerging technologies and approaches, and limited strategic support for newer and untested projects.

These are key insights shared by Angela Hanson, Innovation Lead at OECD, at the third iteration of GovMesh, organized by interweave and held alongside Lithuania’s GovTech Leaders 2025 late last year.

Furthermore, unlike the private sector – where the value of innovation can be measured by financial metrics – the impact of innovation is harder to quantify in the public sector, she noted, which can further stifle efforts to drive innovation internally.

Clarifying strategic intentions

The thing that sets public sector innovation apart, says Hanson, is a focus on improving public good. This can manifest in a variety of ways, from a focus on improving the user-orientation of public services to maintaining the morale of public servants.

Not all of these values can be mapped onto the motivations that traditionally drive innovation within the private sector. This is why it is critical to develop a new way of conceptualizing public sector innovation and the trade-offs that governments need to weigh.

Angela Hanson, Innovation Lead at OECD

In 2018, the OECD’s Observatory of Public Sector Innovation developed the Facets of Innovation framework that aims to support public sector organizations in clarifying the strategic intent driving innovation initiatives, and in turn, better support the longevity of innovation within government.

The framework identifies four types of innovation:

1. Mission-oriented innovation: Achieving an ambitious goal to address a concrete challenge – in the GovTech space, this could look like whole-of-government projects or Moments of Life initiatives.

2. Adaptive innovation: Developing new approaches in light of changing external or internal conditions, such as global crises (e.g., the Covid-19 pandemic).

3. Anticipatory innovation: Building new capabilities that engage with future possibilities, challenges, or emerging needs – for instance, through sandboxes that aim to foster the development of new solutions.

4. Enhancement-oriented innovation: Improving existing systems and practices to achieve better results, which could entail back office modernization or legacy system upgrades.

GovTech as strategic transformation

Hanson points to X-Road as an initiative that encompasses aspects of all four objectives, from its mission to enable whole-of-government services for Estonia to its forward-looking initiative to enable cross-border interoperability with Finland.

In her view, X-Road also serves as a compelling example for the dilemmas that emerge when embarking on innovation. Some of these include:

• Forging interoperability with other countries while preserving sovereignty;

• Standardizing whole-of-government approaches to achieving greater efficiency while managing the risk of long-term dependencies;

• Enabling real-time data exchange while safeguarding data privacy.

“We don’t talk about these trade-offs when we talk about the work that we do, at least not as explicitly as we should,” she says.

Other dilemmas she highlights that leaders commonly face in the GovTech space include:

• Introducing platforms that enable standardization across government while ensuring room for local innovation and experimentation;

• The tension between pilot proliferation and building with scale in mind;

• And procuring solutions purpose-built for government needs while managing costs associated with custom solutions.

This is why it is critical for GovTech leaders to not just occupy operational roles within government but to occupy strategic roles that sit close to the center of government – moving beyond digitalization to whole-of-government digital transformation.

When GovTech leaders occupy these roles, they are better able to examine these trade-offs and make tough decisions.

Approaching new technologies

Clarifying and acting with strategic intent can be a powerful tool for navigating dilemmas and guiding decision-making, particularly when formulating a strategy for adopting new technologies within government.

Today, one of the most pressing challenges that a GovTech leader faces when adapting to technological disruption is knowing when to act, she notes.

This is what’s known as the Collingridge Dilemma: when new technologies emerge, they are easiest to use, adopt, guide, and regulate – but it is often too early to foresee the impact the tool might have. Conversely, when their impacts are better known, it may be too late to change course.

Within government services, the early adoption of new technologies such as artificial intelligence could result in vendor lock-in and path dependency that could inadvertently rack up additional costs in the long run. However, waiting too long to act can result in being left behind.

“Once you’re locked in – and this is a big challenge with AI – when the (contract) renewal comes up in a few years, all of these features are going to cost a lot more money,” she says.

Likewise, new technologies often come with implicit values embedded within them, and a clear minded approach to tech adoption can help to clarify why, how, and where to apply these new tools, she shares.

As countries embark on innovation projects and grapple with internal and external pressures to adopt AI, a structured, intentional approach towards the adoption of new technologies where values are clearly defined can be the first step towards clarifying trade-offs and charting the right innovation strategy.

This week, our focus is back on sovereignty, with news from either side of the European continent. Belgium has released a new government messaging service to wean its civil servants off WhatsApp, while Finland is delaying a move of its citizens’ data to AWS for security purposes.

Our main stories this week:

• Nepal’s stuttering digital identity efforts

• Belgium rolls out a matrix-based messaging app for government

• Finnish parliament delays move of election system to AWS

As a reminder, we have recently begun to expand our digital government coverage away from just Substack and onto Linkedin, Twitter and Instagram. For original content, audience engagement, and roundups of all our pieces, feel free to follow us over on those channels. Also check out our sister newsletter, interweave.africa.

Nepal built a digital identity app. So why do government offices still want the paper? - Aarya Chand, Kathmandu Post

A look from the Kathmandu Post into the disconnect between the ambitions and realities of a digital identity-driven Nepal.

• Despite the launch of the Nagarik ID app and Digital Nepal 2.0 framework, many banks, hospitals and government offices still require citizens to present physical copies of their documents in order to receive services.

• Most of these challenges are driven by weak governance. The Department of Information Technology, for example, does not have enforcement authority when agencies refuse to accept digital IDs.

• Beyond isolated success stories - such as the country’s traffic police - experts say that the rollout will continue to be slow and uneven until Nepal tackles the threefold challenge of legal recognition, institutional mandates and privacy-aware data sharing.

A BEAM of light for EU sovereignty as Belgium rolls out Matrix-based messaging app for government - Paul Sawers, Resilience Media

Belgium has begun rolling out BEAM - a new, secure, and sovereign messaging app for public sector employees and military personnel.

• Belgian Secure Communications (BSC), a federal organization, worked to develop the app within the confines of national infrastructure as a replacement for consumer apps like WhatsApp.

• The app is designed to eventually serve up to 750,000 personnel across all departments and public bodies, providing an end-to-end encryption service for all government messaging.

• It is based on the Matrix protocol - an open-standard for secure, decentralized communication which allows different servers to communicate with one another, much like email.

Finnish government not moving election system to AWS for now - Werner Pluta, Heise Online

Finland has paused plans to move its election processing system to AWS, with next year’s parliamentary elections set to run on previous in-house servers instead.

• The Ministry of Justice had previously wanted to move all data on candidates, voters and vote counting to AWS. But the government is now claiming it is revisiting the situation due to change in global politics and affairs.

• Part of the subtext here is the US Cloud Act, which obliges US companies to hand over data to US authorities regardless of storage location, meaning that the Finnish government could be putting citizens’ data at risk by shifting to the American provider.

GovTech News in Brief

UNDP calls on innovators to build digital tools for crisis impact reporting - UNDP

The UNDP’s second global innovation challenge has launched, calling on startups, researchers and digital developers to build tools that help in crisis reporting.

Japan wants to ‘design’ its way out of bureaucracy to reach its AI-first government ambitions - Si Ying Thian, GovInsider

A look into how Japan is shooting for a complete redesign of the country’s infamous government bureaucracy to meet the needs of the AI era.

Tajikistan manages water with AI innovation - Parvathy Ramaswami, UN SDG

While AI is generally often talked about in a negative light with regard to its relationship with the environment, this UNSDG piece explores how Tajikistan is using AI Innovation for more sustainable development with water resources.

New Zealand to enable digital ID credentials, age checks via Govt.nz app wallet - Masha Borak, Biometric Update

New Zealand is planning to roll out a digital identity wallet by end of March to allow for digital credentials from both the private and public sector, allowing for citizens to have easy access for government services.

New €800,000 Maltese digital platform to bring all local council services together online - Malta Independent

The island of Malta is looking to unify its disparate digital platforms to bring all government services together.

Oman launches digital platform to streamline maritime services - Muscat Daily

Oman has launched a new digital platform to streamline maritime services, eventually covering more than 90 services in a bid to reduce paperwork and simplify procedures for users, companies, and the government.

Bitcoin as public infrastructure: how Bhutan plans to build a city on BTC - Mamsi Nkosi, IT News Africa

Bhutan is positioning Bitcoin as a part of public infrastructure by pledging 10, 000 BTC to Gelephu Mindfulness City, a special economic zone designed to employ digital assets in its financial architecture.

The Theory Behind the Practice

Cities must lead digital shift, says Madrid’s tech chief - Xhoi Zajmi, Euractiv

Madrid’s Digital Office Director makes the argument that cities should be leaders in the digital revolution, arguing that changing culture is a bigger barrier than the technology itself and that cities must work to maintain public trust.