For Hemit HF, the application is seen in the context of the planned collaboration with NorwAI on a project where, together with St. Olavs Hospital, they plan to test out whether NorwAI's newest language model can support the hypothesis that smaller, Norwegian language models, trained and fine-tuned for the health domain, will perform better than large, international generic models.

Born from research

Medbric on the other hand, is a true-born child of research.  From the outset, it grew out of a concrete clinical challenge. In her PhD project at the Antibiotic Center (ASP) at the University of Oslo, physician and researcher Jorunn Thaulows Phd-project showed substantial variation in antibiotic prescribing between doctors- despite limited evidence of benefit in several common scenarios. So far, the measures introduced to reduce antibiotic use have had limited effect, prompting Thaulow to think along new lines: Could artificial intelligence provide solutions to the problem of overuse of antibiotics, she asked.

At NorwAI, Adjunct Associate Professor, Jon Espen Ingvaldsen is part of the team developing Norwegian language models for real-world use. He took Thaulow's findings seriously, and together they developed solutions that meet healthcare requirements for clinical relevance, security, and usability.
- The company was founded on September 17 last year and eight days later Medbric was officially launched on stage during NorwAI's NorwAI Innovate Conference 2024. We build AI agents for the healthcare sector. Our solutions produce precise, personalized clinical documentation, generate summaries and provide quick answers to medical questions, says CEO Jorunn Thaulow.

15 months later

Fifteen months later, Medbric has scaled rapidly. Around 2,000 general practitioners—approximately 25–30% of all GPs in Norway—use the company’s ambient scribe solution. In addition, 1,000 physicians in specialist health services are among the users, along with 30 municipalities. Medbric’s speech-to-note solution, Aila, was launched in September 2025 at 10 hospitals across both the «Helse Vest» and «Helse Midt-Norge» Regional Health Authorities. Medbric has also entered the Danish health sector, adapting the solution to two national electronic health record systems; in Norway, it is integrated with three systems. 

The company is also proud to share that it has recently been awarded NOK 3.4 million from the Research Council of Norway to launch an evaluation study of a new agent for clinical decision support. 

Medbric is built on a foundation of research, and our main focus is staying close to research while delivering quality and trust. We firmly believe that close cooperation with a strong, competent research community is a recipe for us. Another key success factor is a close co-developing with healthcare professionals themselves, says Jorunn Thaulow. 

The tech enterprise

Hemit HF provide IT services for specialized healthcare to the Central Norway Regional Health Authority (Helse Midt-Norge). Hemit HF develops, manages and operates common ICT systems for all hospitals in Trøndelag and Møre og Romsdal. Hemit HF works closely with the hospitals on a joint initiative on AI in the Health Ministry of Mid-Norway, and participates in several national forums dealing with Artificial Intelligence in the health services.

Both prospective partners are excited about the the idea of joining NorwAI. 
- In the specialist health service, many are eager for digital tools that improve their working life. We are anxious to see whether language models adapted to the Norwegian health service can contribute to this, says CEO Trond Utne at Hemit HF.

Why health research matters

NorwAI aims to enter the health domain for research purposes. Professor Eric Monteiro highlighted the challenges when plans for a collaboration with St Olavs Hospital and Hemit HF were announced:

-   It is an exciting research challenge to find good benchmarks for language models within a sector where the tasks vary significantly across fields, from chronic mental health to surgery in all its forms, says Professor Eric Monteiro.

National health authorities have established strict rules for the introduction of AI in the sector. The report: "Use of artificial intelligence in health and care services" (Helsedirektoratet 2025) outlines the key considerations organizations must address to ensure AI systems are trustworthy when procuring, implementing, and using AI in health and care services. 

A rapidly evolving field

The report describes the current legal framework, outlining what is required by applicable regulations and, in part, what is considered good practice. However, AI is a rapidly evolving field, and new legislation is on the horizon. 

In 2019, the European Union published ethical guidelines defining three core components of trustworthy AI that should be upheld throughout a system’s lifecycle: (1) it should be legal, complying with all applicable laws and regulations, (2) it should be ethical, ensuring adherence to ethical principles and values, and (3) it should be robust, both from a technical and social perspective, as AI systems, even with good intentions, can cause unintended harm 
 

By Rolf D. Svendsen

2025-12-17