by Harry Pyant, Enterprise Lead, Healthcare – Telehouse Europe 

As healthcare organisations scale AI, the question is no longer just what the technology can do. It’s about having visibility to understand, manage, and show the infrastructure behind it. 

As AI moves from pilot projects into everyday healthcare operations, healthcare leaders are under growing pressure to prove that innovation is delivering measurable value. In the NHS and across the wider healthcare sector, adoption is taking place against a backdrop of rising demand and constrained budgets. Stretched teams are also being asked to modernise services while showing progress against sustainability commitments. 

That makes AI adoption an infrastructure issue as much as a technology decision. AI can support faster care pathways and help reduce inefficiencies, but it also increases demand on the digital environments behind healthcare services. 

For healthcare leaders, the challenge is becoming clearer. They need to understand how AI is being deployed, how much energy it consumes and whether its environmental impact can be proven. 

For organisations like the NHS – which committed to becoming the world’s first net zero healthcare system in 2020 – this raises an important question: how can healthcare organisations scale AI to support better outcomes while maintaining visibility over the infrastructure that supports it? 

The gap between ambition and evidence 

There is clear momentum behind AI adoption across healthcare. It is being used to support clinical pathways and improve operational efficiency. 

At the same time, boards and governance teams are under growing pressure to ensure sustainability claims can be proved, not assumed. That tension is becoming more visible as AI adoption grows. 

Recent Telehouse research shows how this tension is playing out in practice. Conducted among 500 UK IT decision-makers at major organisations including healthcare providers, our research found that:  

  • 86% of healthcare respondents say AI is helping drive net zero and ESG progress 
  • Only 38% are very confident its efficiency will outweigh the carbon it generates 
  • 56% admit their organisations struggle to measure AI-related workloads accurately 

Taken together, these findings point to a clear gap between ambition and evidence. Healthcare organisations largely see AI as part of the solution, but many still lack the visibility needed to prove its impact or confidently assess the trade-offs behind its deployment. 

What the research shows about investment priorities 

That same gap is reflected in where investment is currently focused. More than half of healthcare respondents (55%) are prioritising skills development, while 54% focus on cost savings, and 48% prioritise sustainability. 

All are important, but sustainability is still not leading the agenda. This matters as physical and digital infrastructure makes up 10 per cent of the NHS Carbon Footprint Plus, a measure that includes emissions across the wider supply chain.  

Yet this is also where many organisations have the least visibility. Without a clear understanding of how infrastructure supports AI workloads – and the energy and emissions associated with it – it becomes significantly harder to measure progress, evidence sustainability claims, or make informed decisions about how AI should be scaled. 

Why infrastructure visibility matters 

For healthcare organisations, the challenge is therefore not simply whether AI can support better outcomes, it is whether the infrastructure behind AI can be managed and measured in a way that supports sustainability goals. 

This is where the underlying data centre environment becomes increasingly important. If AI workloads are hosted on infrastructure where energy use is difficult to isolate, healthcare organisations may struggle to understand the true environmental impact of those applications. That can make sustainability reporting more complex, particularly where emissions sit across different parts of the organisation or its supply chain. 

As AI adoption grows, healthcare leaders will need infrastructure models that provide greater transparency without compromising performance. For many organisations, that means looking beyond where workloads are hosted today and considering environments designed to provide clearer operational and sustainability data. Colocation can support this by giving organisations access to purpose-built environments where energy use is easier to monitor and report. 

This becomes especially important when it comes to energy visibility. If an AI application runs on hospital-owned infrastructure, its electricity consumption may be difficult to isolate unless dedicated metering is in place. Advanced colocation providers can provide clearer visibility of a customer’s energy use within the data centre, where appropriate metering and reporting are in place. That gives healthcare organisations a stronger basis for sustainability reporting and helps them better understand the infrastructure impact of AI as usage grows.  

What healthcare organisations should look for 

Not all colocation environments offer the same level of support for sustainable AI adoption. 

More advanced providers combine advanced liquid cooling technology with guarantees of renewable energy use. They should also offer transparent reporting and metering, so greenhouse gas emissions are fully audited. 

Cooling is a particularly important part of the picture. The computing infrastructure needed to support AI systems is only as reliable as the cooling systems behind it. AI applications drive significantly higher power densities, meaning data centres need more advanced solutions to complement traditional cooling systems. Liquid cooling is set to become more central to these plans because of its efficiency and reduced energy requirements compared with air cooling.  

Supporting more sustainable AI adoption in healthcare 

AI has an important role to play in the future of healthcare. It can help improve service performance and reduce inefficiencies across the system. However, its long-term value will also depend on how sustainably it is deployed. 

For healthcare organisations, sustainable AI adoption starts with understanding the infrastructure behind it. Explore how Telehouse’s colocation environments can support more measurable, efficient, and resilient AI deployment, speak to an expert here.