How healthcare and the environment can benefit from 'Green' AI

Artificial intelligence is becoming deeply embedded in healthcare, from medical imaging and cancer diagnostics to treatment planning, hospital management and telemedicine. These applications can improve efficiency and help clinicians process large volumes of information faster, but they also depend on energy-intensive computing infrastructure. A new research paper published in Artificial Intelligence in Medicine argues that healthcare should begin treating AI’s environmental impact as seriously as issues such as privacy, bias, fairness and transparency.

The idea behind “Green AI” is broader than simply making algorithms more efficient. It asks whether the full environmental footprint of AI, including electricity use, water consumption, data centre infrastructure, hardware manufacturing and end-of-life disposal, is being considered when healthcare systems adopt new technologies.

 

Why AI Has an Environmental Cost

 

Training and operating AI systems requires significant computing power. Much of that activity takes place in large data centres that consume electricity and water, particularly for cooling. As AI use expands, the infrastructure required to support it is also growing.

Alok Mishra, Professor in Data Management and Software Engineering at the Norwegian University of Science and Technology, argues that sustainability needs to become a core part of how AI systems are assessed.

"AI models require lots of energy and huge data centres. These data centres consume large amounts of electricity and use significant quantities of fresh water," Mishra said.

The concern is particularly relevant for healthcare because the sector is adopting AI rapidly. If AI becomes part of routine diagnostics, clinical administration and medical research, its cumulative environmental footprint could increase substantially.

 

Healthcare Has Strong Reasons to Use AI

 

The sustainability challenge is complicated by the clear benefits AI can provide to patients and clinicians. Healthcare professionals increasingly use AI to analyse medical images, support diagnosis, identify treatment options and process large amounts of patient information.

Åsmund Flobak, an oncologist at St. Olav’s Hospital and professor at NTNU, described several practical applications already being used in cancer care. AI is helping researchers analyse living tumour samples, identify how cells respond to drugs and assist clinicians in determining which areas should be irradiated or protected during radiotherapy.

The difficulty is that immediate clinical benefits can outweigh concerns about environmental impact. Flobak said environmental considerations are not typically part of day-to-day decisions when introducing AI into clinical practice, particularly when the technology offers a direct benefit to patients.

 

Read more: ISEP Launches Carbon Accounting Register for Practitioners and Auditors

 

Green AI Requires More Than Efficient Data Centres

 

The research reviewed 47 studies examining sustainability in healthcare AI and found that existing efforts were often fragmented. Individual projects might focus on energy-efficient models or cleaner data centres without considering the wider environmental effects of the technology.

Green AI instead calls for a life-cycle approach. That means examining the environmental consequences of an AI system from the manufacture of computing equipment through development, operation and eventual recycling or disposal.

Under this approach, an AI application would not automatically be considered sustainable simply because it runs on an efficient server or supports a socially beneficial purpose. Its total energy, water, hardware and infrastructure requirements would also need to be considered.

"It is not just a political question anymore. Stakeholders and citizens are facing environmental consequences in daily life," said Mishra.

 

AI Could Also Help Reduce Environmental Pressures

 

The relationship between AI and sustainability is not entirely negative. The same technology responsible for rising electricity demand can also improve the efficiency of energy systems.

The International Energy Agency has estimated that data centre electricity consumption could double by 2030. At the same time, it has identified opportunities for AI to improve the operation of electricity grids, potentially freeing substantial transmission capacity through better forecasting and system management.

Healthcare could face a similar balance. AI may increase the sector’s digital footprint, but it could also reduce waste, improve resource allocation, accelerate research and help clinicians deliver more targeted treatments.

The challenge is determining whether those benefits outweigh the environmental resources required to achieve them.

 

Explore OneStop ESG Marketplace: Environmental monitoring & sensing

 

Healthcare May Need New Sustainability Metrics for AI

 

One implication of Green AI is that healthcare organisations may eventually need clearer ways to compare the environmental performance of different AI systems. Mishra compares the idea to energy-efficiency ratings on household appliances or carbon information provided by airlines.

That would require greater transparency from AI developers and data centre operators about electricity consumption, water use and computing requirements. The research acknowledges that this could be difficult because technology companies do not always disclose detailed information about their infrastructure or energy consumption.

Without better data, healthcare providers may find it difficult to assess whether one AI system delivers the same clinical benefit with a lower environmental footprint than another.

 

The Goal Is Better AI, Not Less AI

 

The Green AI concept does not argue that healthcare should stop adopting artificial intelligence. Instead, it asks developers, policymakers and healthcare organisations to consider environmental performance when deciding how AI should be designed and deployed.

"Environmental consciousness therefore has to be at the core of AI development and application in healthcare," Mishra said. "AI use and its applications should be part of the picture, but they should also take into consideration sustainability issues."

As AI becomes more important to medicine, the sustainability question is likely to become harder to separate from clinical and economic decisions. The long-term challenge is to retain the benefits of AI while reducing the energy, water and infrastructure demands required to deliver them.