The Ministry of Digital Affairs Grants a Total of 30.7 Million DKK for an Ambitious Danish Language Model Project

Instead of competing directly with global tech giants like Google and OpenAI, Denmark has chosen to pool its resources to address well-defined tasks. A unified national effort aims to make a significant impact.

This approach ensures that Danish investments in AI are directed toward solutions that meet critical and specific societal needs while promoting sustainable development and a fair digital economy.

The Ministry of Digital Affairs has granted a total of 30.7 million DKK to the project as part of the government’s strategic focus on artificial intelligence. The funding will support testing of various use cases involving public administration, the education and healthcare sectors, as well as small and medium-sized enterprises.

– By integrating stringent security protocols with collaborative and user-driven flexibility, DFM aims to fully harness the potential of AI to meet Denmark's diverse societal needs and pave the way for a better future. We aspire to bring technology and human expertise together for the common good, explains Peter Schneider-Kamp, Professor at SDU’s Department of Mathematics and Computer Science, Data Science Section.

The primary goal of the project is to establish a secure R&D platform for training, fine-tuning, evaluating, and maintaining foundation models for Danish-language applications. This platform will adhere to the highest standards of data integrity and documentation for safe AI use.

Additionally, an innovative and open "sandbox" environment will facilitate ongoing collaboration for the fine-tuning and adaptation of foundation models. Domain experts from national projects will collaborate in this secure and flexible space to design and improve specific use cases.

– DFM is built on the vision that community and inclusion must guide the development of Danish language technology. Our interactive sandbox will bring together researchers, developers, and users to rapidly and flexibly create prototypes and refine solutions for various societal needs, says Kristoffer Nielbo, Professor at Aarhus University’s Center for Humanities Computing.

– Drawing on Denmark’s cultural heritage and tailoring existing models with cultural sensitivity, we aim to create language technology that reflects and respects the complexity of Danish society—past and present. This commitment to openness and transparency underpins our research and development, not only to deliver the best models but also to provide tools and documentation for further development and reproducibility, Nielbo adds.

The platform will leverage existing collaborations, supercomputing infrastructure, and data-sharing agreements to significantly reduce establishment costs. The models, both foundational and fine-tuned, will be made available as open-source tools via the R&D platform for both public and commercial use.

– Developing language models is a long and challenging process. We started this work many years ago, and with this new funding, we can make the necessary progress together. As a GTS institute, it is precisely our role to bridge research and business. This funding is critical for continuing our strong collaboration, enabling us to train, fine-tune, and maintain Danish foundation models. Ultimately, this means sustainable AI solutions based on applied research and real use cases for Denmark, emphasizes Jens Kaas Benner, Head of the AI Lab at the Alexandra Institute.

– Within the consortium, we will ensure that the language models are evaluated in the cultural and societal contexts they are intended for. This will be done through a series of Danish-based benchmarks that assess the models' general language understanding and incorporate knowledge about Danish language and culture, adds Bolette Sandford Pedersen, Professor and Deputy Head of Department at the University of Copenhagen.

Further information

For further information, please contact: Peter Schneider-Kamp, Professor, Department of Mathematics and Computer Science, Data Science Section, SDU, Phone: +45 6550 2327, Email: petersk@imada.sdu.dk or

Jens Kaas Benner, Head of AI Lab, Alexandra Institute
Phone: +45 2091 1103, Email: jens.benner@alexandra.dk