The 2025 Europe edition of the Data for Policy conference, held 12–13 June at Leiden University in The Hague, gathered researchers, policymakers, and practitioners around the theme “Twin Transitions in Data and Policy for a Sustainable and Inclusive Future.” Over two days of presentations and discussion, participants explored how data systems and digital technologies are reshaping governance and what it will take to make them serve climate goals, social equity, and democratic accountability.

Here are five reflections from the event.

1. Europe Wants to Align Digital and Green Goals, but the World Isn’t All In.

Andrea Halmos of the European Commission opened the conference with a call to integrate Europe’s digital and environmental agendas. She argued that digital tools and sustainability can be mutually reinforcing, supporting emissions tracking, predictive resource management, and real-time environmental monitoring. She urged public institutions to lead this integration, pointing to initiatives like climate-neutral data centers and digital twins for monitoring circular economy efforts.

Halmos warned, however, that local governments lack the technical capacity to implement these goals. Municipalities are missing up to 70 percent of the skills needed to meet EU climate and innovation targets. Her keynote also highlighted an underlying tension: green innovation depends on digital infrastructure, but that infrastructure through energy use, material demands, and surveillance risk can undermine climate commitments. The EU is developing initiatives such as the Cloud and AI Development Act and sustainable digital standards to manage these tradeoffs, but Halmos emphasized that implementation will lag without stronger coordination and investment in local operational capability.

A comparative study by researchers at the University of Copenhagen added an international perspective. Analyzing how Europe, the U.S., and China frame “green IoT” in policy, they found that only Europe treats digital and environmental goals as interlinked. U.S. policy focuses on market growth and national security, while China emphasizes industrial expansion and platform integration. These divergent priorities shape how each region approaches tech development, regulation, and collaboration.

2. AI and Digital Systems Must Be Built Around Justice Instead of Efficiency

Several sessions pushed back on the idea that digital innovation alone will deliver a just and sustainable future. Instead, they asked whether today’s AI and data systems are even compatible with those goals.

Researchers from University College Dublin analyzed how EU and national policies fail to account for the environmental impacts of AI infrastructure. Focusing on Ireland’s data center boom, they showed that hyperscale facilities, essential to AI development, consume nearly a quarter of the country’s electricity while operating with limited transparency. Regulatory blind spots and trade secrecy claims, they argued, sustain an artificial separation between AI applications and the infrastructure that supports them. This disconnect hampers governance and obscures the true operational costs of emerging technologies.

Barrie Sander of Leiden University extended this critique, describing AI as an active participant in climate governance: a resource consumer, a tool for mitigation and adaptation, an enabler of securitized responses to climate migration, and a shaper of narratives through corporate lobbying. Rights-based regulation, he argued, may protect individuals, but it falls short against systemic risks and planetary boundaries. What’s needed is a shift that enables legal frameworks to match the scale and systemic logic of industrial AI.

Researchers from the Tilburg Institute of Law, Technology & Society added a sectoral view, focusing on the EU’s clean energy transition. Their work showed how digitalization is framed in technocratic terms optimizing for competitiveness or system efficiency , while sidelining civic democratic values like equity and justice. Market-based tools such as tariff adjustments or digital consumer dashboards cannot address deeper problems like energy poverty or infrastructure inequality. They called for a re-framing of the twin transitions that centers inclusion and justice rather than optimization alone.

3. Data Governance Aims High, But Stumbles on the Ground

Europe’s data policies often seek to promote the public good. Yet multiple presentations showed how those same frameworks can stall sustainability innovation, particularly in cases that depend on using personal or biometric data.

Andres Chomczyk Penedo and Anna Capellà Ricart examined how GDPR complicates the development of extended reality (XR) tools for green innovation. These tools can support urban planning or energy management, but regulatory uncertainty, limited infrastructure for managing consent, and privacy-by-design constraints restrict experimentation and slow deployment.

Adrianna Michałowicz offered a similar critique of the EU’s Data Governance Act. While it encourages voluntary data sharing through mechanisms like data altruism, she found that participation has been low. Weak incentives, vague definitions, and a lack of institutional support have left the concept underdeveloped. Unless trust and accountability mechanisms improve, these policy tools may struggle to deliver.

4. Decentralized Models Offer a Way Forward

Despite these regulatory barriers, several presentations showcased alternatives that put privacy, participation, and sustainability at the center of digital design.

Researchers from Dublin City University shared insights from the SHINE project, which uses indoor air quality sensors in social housing. To overcome concerns about data privacy and loss of control, the team developed a sensor system that embeds transparency and community control directly into its design. Privacy was not treated as an afterthought but as a foundational principle, shaping both the technical architecture and how residents engage with the system.

At the London School of Economics, researchers explored person-centric data architectures like Personal Data Stores. These tools allow individuals to manage how their data is shared, offering an alternative to corporate surveillance and centralized state data collection and creating space for data practices that are granular, participatory, and accountable. Their project examined how these systems could contribute to national statistics and circular economy metrics while allowing people to choose what data they share, with whom, and for what purposes.

Adding an institutional dimension, Federico Bartolomucci shared research performed with co-authors on the MIT City Science Network. This network links urban innovation labs across multiple countries, each serving as a testbed for participatory data practices shaped by local needs. Through interviews and comparative analysis, the researchers found that these labs succeed not by standardizing approaches but by fostering adaptable governance, shared learning, and community-led experimentation. The result is a decentralized model of innovation that prioritizes collective outcomes without imposing uniform solutions.

5. Public Institutions Need Better Ways to Access Private Data

Many of the data sources most relevant to public policy ,on mobility, consumption, platform use are held by companies. Governments need this data to make informed decisions, but access remains limited, fragmented, and often informal.

A dedicated track at the conference tackled this issue. Researchers from Utrecht University and the University of Manchester analyzed responses to the EU’s public consultation on the Data Act. They found a sharp divide: public actors favored mandatory sharing for public interest purposes, while private firms favored voluntary sharing, safeguards, and price-setting mechanisms. While both sides acknowledged the problem of asymmetrical access, there was no consensus on how to address it.

Another Utrecht team examined B2G dynamics in Amsterdam and Barcelona. Drawing on discourse analysis, they found that trust gaps and unclear legal frameworks often stall collaboration. While governments try to insert data-sharing clauses into procurement and permitting, companies remain cautious without clear incentives.

The GovLab offered a possible bridge. Adam Zable presented the “9Rs Framework,” a taxonomy of business motivations for sharing data, including Reputation, Reciprocity, Regulatory compliance, and Revenue generation. The framework offers a structured way to identify and surface what drives companies to share data, aiming to help policymakers and private organizations design data collaboration initiatives that benefit all parties. Zable outlined plans to test and refine the framework through interviews with corporate data stewards.

Conclusion

Data for Policy 2025 made clear that aligning digital and green transitions will require not only deploying new tools but, more crucially, rethinking the infrastructures, incentives, and institutions that govern how data is produced, accessed, and used. Across sectors, presenters emphasized that a deep shift is needed toward data practices that are participatory, transparent, and accountable to the communities they affect. For the twin transitions to succeed, governance must be shared, values clearly stated, and technology anchored in public purpose.