Leap Trends · Technology · Adoption
Briefing
AI use, training and AI policy among healthcare professionals
- Healthcare
- Training
- AI policy
More than half of nurses, carers and support workers use AI, mostly on their own initiative and without training. Hospitals are adopting AI policies at speed, but half of healthcare professionals see no clear policy and two thirds feel insufficiently prepared.
This briefing has not been updated for more than three months. Check the figures against the source before you use them.
Compiled with AI from public sources, with the source for every figure. Key figures labelled Checked have been compared with the source by a person. How that works
Key figures
- 54%of nurses, carers and support workers used AI at work at the end of 2025; two thirds of them started on their own initiative.NivelNot yet checked
- almost 90%of these healthcare professionals have never had any training in AI; more than six in ten would like some.NivelNot yet checked
- 76%of hospitals have an approved AI policy, against 50% in 2025 and 30% in 2024.M&I/PartnersNot yet checked
- 21%of hospitals train staff in using their own AI tools; another 21% have no active policy on it.M&I/PartnersNot yet checked
What the sources say
In September 2025 Nivel, the Netherlands Institute for Health Services Research, sent an online questionnaire to its national Nursing and Care Panel, supplemented by an open link shared via the nurses’ association V&VN, LinkedIn and its own website. Responses came from 890 nurses, carers, nurse practitioners, support workers and practice nurses working in hospitals, mental health care, disability care, community nursing, GP practices and elderly care; the results were weighted by sector and level of education and published in April 2026. Just over half (54 per cent) used AI at work, ranging from 85 per cent of nurse practitioners to 38 per cent of carers with a vocational qualification. Most often it is used to look up information; 16 per cent use AI for record-keeping in the electronic client file. Seven in ten users have been doing so for less than a year, and for two thirds the initiative was their own.
These professionals see less of what their own organisation is doing. A fifth feel that their organisation actively encourages the use of AI. Half miss a clear organisation-wide AI policy, most often in GP practices, and say they are not involved in decisions about the use of AI. Three in ten know of ethical guidelines for AI in their organisation. Almost nine in ten have never had any training in AI and two thirds feel insufficiently prepared. At the same time, two thirds believe that they themselves are responsible if AI makes a mistake.
In March 2026 M&I/Partners published the eighth AI Monitor Ziekenhuizen, its monitor of AI in Dutch hospitals, with responses from 52 hospitals (73 per cent of the total). The 56 respondents are mainly CIOs, programme managers and clinical informaticians. In those hospitals policy grew sharply in a year: 76 per cent have an approved AI policy, against 50 per cent in 2025 and 30 per cent in 2024, and 60 per cent have an AI board or steering group. Generative AI is used in 90 per cent of hospitals. For the AI tools staff bring in themselves, 54 per cent have ground rules, 29 per cent offer their own secure AI environment and 21 per cent provide training. Almost all hospitals are working on AI literacy, mainly through communication, guidelines and e-learning; 21 per cent have an organisation-wide training plan. Staff awareness and knowledge of AI is named by 42 per cent as an area that is not in order.
On 6 October 2025 the Dutch Minister of Health, Welfare and Sport announced the Programme for Realising AI in Healthcare, with the ministry in a coordinating role and, as agreed in the AZWA (the supplementary health and social care agreement), an initial focus on reducing paperwork. The letter places final responsibility with the healthcare professional who uses the application. The two surveys look at different groups: hospital staff in the monitor, frontline professionals in six sectors at Nivel. The two groups cannot therefore be compared like for like. Set side by side, though, the framework in hospitals is growing faster than the training of the people who, according to the minister, carry the responsibility.
How solid is this? The Nivel survey is the first of this size among nurses and carers and captures a single moment, the end of 2025. The hospital monitor does have a series, but the answers come from management and IT. Only a further measurement will show whether the gap between policy and the front line is narrowing.
For the leadership team
An approved AI policy is not the same as a carer who knows what she may do with an AI-generated summary in the patient record. The question for the board is how the organisation knows what staff are already doing with AI, and who teaches them to check the output if they remain responsible for the mistakes themselves.
Sources
- Nivel · Aafke Arts, Kim de Groot, Maaike Olivier, Sofie Noorland and Anneke FranckeKunstmatige intelligentie (AI) in de zorg. De meningen van verpleegkundigen, verzorgenden, verpleegkundig specialisten, begeleiders en praktijkondersteuners
Bijna negen op de tien respondenten rapporteren dat zij nooit scholing over AI hebben gevolgd.
- M&I/Partners · Wouter Gude et al.AI Monitor Ziekenhuizen 2026. Wat is de stand van zaken van AI bij Nederlandse ziekenhuizen?
Ziekenhuizen laten zien dat het opstellen van kaders, normen en werkafspraken relatief snel gerealiseerd kan worden.
- Ministerie van VWSBrief van de minister van VWS over de inzet van AI in zorg en welzijn (Kamerstuk 27 529, nr. 350)
AI is hierbij nadrukkelijk ter ondersteuning, de eindverantwoordelijkheid ligt bij de zorgprofessional die de AI toepassing gebruikt.
We summarise in our own words and quote no more than one sentence per source. Read the original: there is more in it. How we use sources.
Academic research
- Halmstad University · Petersson et al.Shadow AI in Swedish Health Care: Qualitative Analysis of Physicians’ Free-Text Answers
- Hellenic Open University · Kontodimopoulos et al.Digital Skills and Readiness of Greek Nurses for Artificial Intelligence Adoption in Clinical Nursing Practice
- California State University, Dominguez Hills · Oleribe et al.Healthcare Providers’ Perspectives on Generative Artificial Intelligence (GenAI) Adoption, Adaptation, Assimilation, and Use in the United States
Academic articles on the same phenomenon, no more than a year old at the last search, from universities in Europe and North America, found through the open database OpenAlex. These articles count towards the publications per month; the text and key figures above are based on the sources. How we choose the articles.
How this briefing has been updated
Title now describes the topics; previously “Healthcare professionals already use AI; training and policy come later” (AI)
Added 3 articles from universities in Europe and North America (via OpenAlex). (AI)
First draft based on the Nivel survey of 890 healthcare professionals, the AI Monitor Ziekenhuizen 2026 (AI monitor of Dutch hospitals) and the letter to parliament on AI in healthcare. The AMA survey of US doctors is not included because the page could not be opened. (AI)