This focus area is based on results from the knowledge base on large language models, which provides an overview of the use and development of large language models in the health and care sector in Norway. The report describes how such models can help to improve healthcare, including benefits and challenges, and identifies possible measures to meet the challenges and strengthen the use of language models in the sector (section 5.1).
Two measures were assigned to the Norwegian Directorate of Health in the 2024 letter of allocation: Assess the risks associated with the use of large language models (section 5.2) and facilitate the use of language models adapted to Norwegian conditions (section 5.3).
The other recommended areas concern the strengthening of expertise relating to the appropriate use of large language models and the highlighting of areas where large language models are suitable and can offer benefits in the form of more efficient work processes or better quality services in the Norwegian health sector. To better understand the needs and measures that will be most appropriate and who should take responsibility for them, these topics will be addressed in the seminar series on the integrated use of AI in the health and care services (section 1.2).
Background
In 2024, the Norwegian Directorate of Health delivered a knowledge base on large language models in the health and care services. The report recommends measures in the following areas:
- Risk assessment
- Secure access to language models adapted to Norwegian conditions
- Strengthening of expertise concerning use, development and fine-tuning
- Assessment of applications for large language models
The recommendations from the report have been incorporated and further developed in the description of the measures below.