Positively Green

Artificial Intelligence (AI) is quickly becoming part of everyday life in the NHS. Whether it's drafting emails, summarising documents or helping with research, more and more colleagues are using AI tools as part of their daily work. At University Hospitals Plymouth NHS Trust (UHP) alone, more than 575,000 AI interactions were recorded in just 14 days, equivalent to around 15 million interactions a year if usage continues at current levels.

 

It's easy to understand why. AI can help reduce repetitive administrative tasks, save time and improve productivity. For some colleagues, including those with dyslexia and other neurodivergent conditions, it can also make writing and processing information more accessible.

 

As AI becomes more embedded in the way we work, it's worth taking a moment to consider something that often goes unseen: the resources needed to power it. Despite the term "cloud", AI isn't floating somewhere in cyberspace. Every prompt relies on physical infrastructure. Behind the scenes, large data centres filled with powerful servers process vast amounts of information, requiring significant amounts of electricity to run and water to keep them cool. At current levels of use, AI activity across UHP is estimated to require around 43,000 kWh of electricity each year, equivalent to powering approximately 14 average UK homes for a year. While few of us will ever see the data centres supporting these technologies, they remind us that digital tools have a real environmental footprint through their use of both energy and water.

 

At first glance, this might seem like an argument for using less AI. However, the picture is more complicated than that. Healthcare also has a significant environmental impact. Travel, paper-based processes, duplication of effort and inefficient use of resources all contribute to emissions. If AI can help streamline administrative tasks, optimise scheduling, support virtual care or reduce unnecessary appointments, it may also help reduce some of these impacts. Research also suggests AI has the potential to support improvements in healthcare and more efficient use of resources when implemented responsibly.

 

Perhaps, then, the question isn't whether AI is good or bad. The reality is likely somewhere in between. Like any technology, AI brings both opportunities and risks. It has the potential to improve efficiency and accessibility, but it also consumes energy and water that should not be ignored.

 

Equally important is recognising what AI cannot do. AI should support human expertise and judgement, not replace it. While it can help colleagues find information, summarise documents and reduce administrative burden, it cannot replicate professional experience, critical thinking or the human connection that sits at the heart of healthcare.

 

AI may also become an important tool in supporting sustainability itself. As healthcare systems become increasingly complex, AI has the potential to help organisations better understand how resources are used and where improvements can be made. Some technology providers, including Microsoft, describe AI as an enabler for improving efficiency, sustainability measurement and environmental decision-making, while also recognising that AI itself consumes energy and resources. This could mean identifying opportunities to reduce waste, predicting demand more accurately and helping organisations make more informed decisions about how resources are used. However, these potential benefits need to be balanced against the energy and water required to power AI systems.

 

As AI becomes an increasingly common part of working life, perhaps the most important question to ask is not "Should I use AI?" but "Is using AI creating enough value to justify the resources it consumes?" That's where the sustainability conversation really begins.