Navigating the regulatory space for MedTech AI

Artificial Intelligence (AI) holds unprecedented potential to transform how healthcare is delivered and will be vital to enable qualitative and cost-effective care in the future. However, the regulatory landscape for the intersection between healthcare and AI creates new challenges for innovators. Challenges that need to be successfully managed to launch new AI-medtech.

Healthcare needs AI to face tomorrow’s challenges

As the world population grows and life expectancy increases, healthcare needs technological advancements to be better equipped for future patients needs. In parallel, vast amounts of data is generated within the healthcare underpinned by rapid digitalization, sensor use and multiomics advancements. AI and Machine Learning (ML) have substantial potential to make sense of complex and large data sets and thereby contribute to more precise, accurate and efficient diagnoses and treatment.

MDR most significant regulatory change in 20+ years

The MedTech industry is heavily regulated to ensure the safety of products released to the market. In the EU, the Medical Device Regulation (MDR) was put into force last year, imposing a major increase of the requirements on MedTech companies. For In-Vitro Diagnostics, the new regulation (IVDR) will be effective from May 2022. However, as of September 2021, only 31 IVDR certificates had been issued1. The Notified Body (NB) capacity has been, and still is, a large bottleneck for market clearance. Historically, starting the transition early and having close communication with NB has been a key success factor for companies in their transition.

EU working on separate legislation to cover AI and ML

While MDR is 174 pages long, it does not cover regulations for AI devices. Instead, EU published an omni-industry proposal called Artificial Intelligence Act (AIA) in April 2021. In this proposal, devices covered under MDR or IVDR will automatically be classified as high-risk AI devices and pursuant to applicable requirements in AIA2. The proposal received critical voices from both industry and NBs since it is highly overlapping with existing regulations. Team-NB, the European Association of Notified Bodies active in the Medical device sector, posted a position paper to express their concern in June 2021. The proposal stated that one major challenge for all actors will be to attract AI experts to build up the expertise needed for the conformity assessment of AI devices3. Thus, the already limited NB capacity will likely be even worse for AI medical devices.

Adaption needed to comply in a regulatory jungle

Moving into a reality where more and more products will incorporate components of AI, MedTech companies will not only have to gain the technical knowledge but also the specific regulatory knowledge. Solely transitioning to MDR is costly for organizations, introducing additional frameworks on top that could potentially pose an obstacle for the continued  development of AI in MedTech. As it stands, the regulatory jungle for AI medical devices is growing, and organizations need to adapt to comply.

What can MedTech AI companies do already now?

Today, the regulatory landscape for AI devices in MedTech is not clearly defined, making it difficult for companies to understand the exact requirements. Taking a holistic perspective and focus on how to develop a safe product, both from a clinical and data perspective will be a good start. The development and launch of AI devices, will require organization wide continuous learning. More specifically the regulatory department needs to be involved from the start, gaining both product and domain knowledge. Doing so, the regulatory efforts can be accelerated once the application of AIA are approaching and, hopefully, most of the ambiguity has been dispersed. Additionally, the NB bottleneck needs to be considered, and communication set up early. To conclude, just as for MDR, the first movers will have a significant advantage when it comes to launching AI enabled devices and the certification thereof.

Note: A notified body is an organization designated by an EU country to assess the conformity of certain products before being placed on the market. These bodies carry out tasks related to conformity assessment procedures set out in the applicable legislation, when a third party is required4.


1 Ola Philipson for Swedish Medical Products Agency. IVDR Summit. 2021-11-18 
2 European Commission. Laying down harmonised rules on Artificial Intelligence (Artificial Intelligence Act) and amending certain Union Legislative Acts. 2021-04-21
3 Team NB. Team-NB Position Paper. 2021-10-06
4 European Commission. Notified bodies. Retrieved from 2021-01-19

Inställningar för cookies
Vår webbsida använder cookies för att hemsidan och våra tjänster ska fungera som de ska. Cookies hjälper oss att förbättra användarvänligheten för dig som besökare, och därför är vissa cookies nödvändiga för att webbsidan ska vara fullt fungerande. Nedan kan du läsa mer om våra cookies och vilka som är valbara.
Inställningar för samtycke
Inställningar för samtycke
Nödvändiga cookies
Dessa cookies används då våra besökare använder en funktion på vår webbplats och är nödvändiga för att webbplatsen ska fungera fullt ut. Dessa cookies kan därför inte stängas av. Cookies används exempelvis då du fyller i ett formulär eller skapar ett konto och lagrar ingen personlig identifierbar information.
Prestanda cookies
Denna typ av cookie hjälper oss att följa antal besök på vår webbsida och hur våra användare hittade till oss. Vi använder cookies till att mäta och analysera för att exempelvis kunna förbättra användarvänligheten på vår webbplats. Vi kan med hjälp av våra insamlade cookies analysera hur användaren navigerar på webbplatsen, och ta bort irrelevanta sidor och information för att skapa ett så användbart material för våra kunder som möjligt. Den information som vi sparar är exempelvis vilka produktsidor som du besöker.
Dessa cookies används för att kunna analysera hur vi kan marknadsföra våra produkter och tjänster. Insamlingen av cookies kan exempelvis hjälpa oss att anpassa annonser till våra besökare baserat på dess tidigare användning av våra tjänster.