Denmark’s Chamber of Commerce, Dansk Erhverv, is leading a national consortium of public and private organizations to introduce a Danish language version of ChatGPT.
The Danish Language Model Consortium (DLMC) will offer a responsible AI model that will be free for “principled” individuals and organizations in Denmark. Led by Dansk Erhverv, IBM Denmark, and the Alexandria Institute, the project aims to develop a customizable Danish language model to cater to the specific needs of businesses and private users.
In addition to creating a Danish language model, the project will also include a chatbot to assist with everyday tasks such as tax filing. Partners in the DLMC have committed to ensuring transparency and copyright protection for Danish data to ensure safe and ethical use of the technology.
The project is working to align with Danish law and EU regulations, particularly GDPR, and aims to unveil the first Danish language models in late 2024. Casper Klynge, former deputy director of Dansk Erhverv, has joined Zscaler in
San Jose, California, to work on digital transformation.
The DLMC project gained government support in 2023 to develop a Danish language chatbot, with a focus on using data for training Danish language models. The Danish Foundation Models (DFM), a collaboration between the Alexandria Institute and three universities, will play a critical role in the project.
The project is seen as a national asset for Denmark’s digital development, with a focus on AI and natural language processing. The consortium aims to position the Danish language model as a competitive tool in the digital landscape.
The DLMC project has attracted over 20 partners from various sectors, including finance, technology, and emergency services. While the technology will be open source, measures are in place to protect the methods and data configurations used in creating the language model. Topdanmark, a founding partner, plans to implement the Danish language model in its financial services business.
