The Swedish government has officially recognized data as a critical national resource, designating it as “the gold of our time” in a major new directive. In a press release issued in late January 2026, Civil Affairs Minister Erik Slottner announced a formal mandate for Digg (the Agency for Digital Government) to lead a national effort to increase data accessibility and accelerate AI development. This move is a cornerstone of Sweden’s broader Digitalization Strategy, aimed at making the country a global leader in the digital transition by 2030.
For the data and AI community, the focus is squarely on the creation of a “national collaboration structure.” The government is pushing for the establishment of secure processing environments that allow data to be treated and analyzed without compromising privacy or control. By fostering these environments, the state aims to lower the barriers for both public and private actors to build and train AI models using high-value public sector datasets. This initiative isn’t just about technical infrastructure; it is about creating a unified, legal, and secure pathway for data sharing across municipalities and agencies, effectively turning fragmented silos into a cohesive engine for innovation.
Announced on January 23, 2026, this government-led directive aims to centralize the coordination of data sharing and AI implementation across Sweden’s public sector to address the primary bottleneck for regional AI development: access to high-quality, localized data.
By tasking Digg with the oversight of these new “data highways,” the government signals a shift toward a more centralized digital governance model intended to spark a surge in public-private partnerships and accelerate the deployment of AI-driven public services. The strategic core of this pivot is a move toward Federated Data models, where the government backs “secure treatment zones” that allow data to remain in situ while AI models are trained on-site. This pragmatic approach is being hailed by practitioners as a vital way to bypass the typical GDPR-induced paralysis that often stalls large-scale public sector tech projects, ensuring that “intelligence” comes to the data rather than moving sensitive information into risky central repositories.
With the government now officially treating data access as a matter of national competitiveness, the “data silos” of the past are finally being dismantled. The question for the community now is: how quickly can we build the models to match the scale of the data being unlocked?
