AI in Healthcare: Three Business Models That Could Transform Healthcare
- Deborah Nas
- Jan 9
- 4 min read
Artificial intelligence is rapidly reshaping healthcare. Large Language Models (LLMs) like ChatGPT are moving beyond simple chat interfaces and are increasingly embedded in healthcare workflows, patient support, and clinical systems.
In a recent conversation on BNR Nieuwsradio, I discussed three emerging business models for AI in healthcare. Each of them addresses a different part of the healthcare system: patients, hospitals and doctors.
Together they illustrate a bigger shift: AI is not just a tool, but a new layer of infrastructure in healthcare.
1. AI as a personal health companion
A few days ago, OpenAI announced ChatGPT Health, a dedicated health environment inside ChatGPT.
The idea is simple but powerful: a 24/7 personal health assistant that helps people understand and manage their health. Instead of generic internet searches, the system can connect to personal data sources such as:
Apple Health
MyFitnessPal
Electronic medical records
This means users could ask questions like:
“What does my cholesterol level mean for my diet?”
“Prepare me for my appointment with the cardiologist tomorrow.”
“What patterns do you see in my sleep and heart-rate data?”
In other words, the AI becomes a health companion that helps patients interpret their own data and navigate the healthcare system.
OpenAI placed this functionality in a separate health tab, with its own memory and additional privacy safeguards. According to the company, health conversations will not be used to train the underlying models.
I remain somewhat sceptical about such promises. But the concept itself addresses a real need: many people struggle to interpret their health data and often turn to Google or AI tools anyway.
The business model
The expected business model is Direct-to-Consumer.
Patients will likely pay for access through a subscription model. Initially, OpenAI may offer it within the free version to gather usage data and improve the product, but in the long term it is likely to become a premium feature within ChatGPT subscriptions.
However, European users will probably have to wait. Because of strict healthcare regulations and privacy laws, it may take quite some time before this functionality becomes available in Europe.
2. AI infrastructure for hospitals
While the first model focuses on patients, the second model focuses on healthcare organisations. Here, the key players are big tech companies providing AI infrastructure for medical applications.
A good example is Google with Med-PaLM 2 and its broader MedLM suite. In this model, hospitals and health-tech companies use AI models as building blocks to develop their own applications. For instance, Mayo Clinic uses these technologies to search through vast collections of internal medical records.
Doctors often spend significant time digging through patient files to find specific lab results or notes from earlier treatments. AI can act as an intelligent search layer across millions of records, instantly retrieving relevant information.
The goal here is not patient interaction but efficiency: reducing administrative workload and allowing medical professionals to focus more on patient care.
The business model
This follows a classic Infrastructure-as-a-Service (IaaS) model. Healthcare organisations pay for:
usage of AI models (often measured in tokens)
data storage
computing resources in cloud environments
In other words: hospitals pay for access to an AI factory that powers their own digital healthcare tools.
3. AI assistants for doctors
The third model focuses on the daily work of doctors, especially general practitioners.
A great example from the Netherlands is Juvoly. Juvoly developed an AI system specialised in Dutch speech recognition, including the many dialects spoken across the country.
Their technology listens during a consultation between a doctor and a patient. The conversation is automatically transcribed and then summarised using another AI model into the medical SOEP structure used in patient records:
Subjective – what the patient reports
Objective – observations and measurements
Evaluation – the doctor’s assessment
Plan – next steps in treatment
For doctors, this means dramatically less time spent on administrative tasks such as typing consultation notes.
The business model
Juvoly uses a software licensing model targeted at general practitioners. Practices can pay:
a monthly fee per user
or a fee per registered patient
The time saved on administration often offsets the subscription cost. Additionally, Juvoly offers its technology through an API, allowing other healthcare software providers to integrate the speech-recognition capabilities into their own systems.
A broader shift within AI in healthcare
These three examples show how AI is entering healthcare through different entry points:
The personal health companion for patients
The AI infrastructure layer for hospitals
The digital assistant for doctors
Despite the differences, the underlying theme is the same: healthcare systems worldwide face increasing pressure due to ageing populations, rising costs and staff shortages.
AI offers tools that could help keep the system sustainable: by improving efficiency, empowering patients and reducing administrative workload.
At the same time, these developments raise important questions about data ownership, privacy and governance. The organisations that control the data and the infrastructure will likely play a key role in shaping the future of healthcare.
For Europe in particular, the challenge will be to maintain strategic autonomy and ensure that innovation does not come at the cost of control over sensitive health data.
More information?
I am a keynote speaker on artificial intelligence, emerging technologies and the future impact of innovation. As a professor at Delft University of Technology, I study how new technologies move from research labs into real-world applications and what that means for organisations and society.
In my work, I translate complex technological developments into clear insights and practical implications for leaders and decision-makers. I regularly share my perspective in the media, including on BNR Nieuwsradio, where I discuss the latest developments in AI, technology and new business models.
Through my keynotes, research and media work, I help organisations understand not just what new technologies can do, but what they will mean for their strategy, their people and their future.
