Andrew Bowdler, Principal Data Analyst, MIAA writes, Recent reports suggest NHS England will prioritise emergency departments and outpatient services for the first wave of centrally funded Ambient Voice Technology (AVT) implementation, with wider deployment across other sectors expected to follow as national programmes mature.

At the same time, the Medicines and Healthcare products Regulatory Agency (MHRA) has published new guidance clarifying how AVT products should be classified under existing UK medical device regulations. The guidance provides an important distinction between technologies that primarily support documentation and administrative workflows, and those that directly influence diagnosis, treatment decisions or automated clinical actions. 

The NHS's AI journey

While this is an important regulatory development in its own right, the wider significance may be what it tells us about the NHS's AI journey.

For several years, many organisations have explored artificial intelligence through pilots, proof-of-concepts and innovation programmes. Today, the conversation is changing. AI is increasingly moving beyond experimentation and becoming embedded within operational clinical and corporate services. This transition from innovation activity to business-as-usual deployment represents a significant moment for healthcare organisations. 

Technologies such as Ambient Voice Technology have the potential to reduce administrative burden, improve the clinical experience and release more time for patient care. Across the wider NHS landscape, AI is also being explored to support operational planning, patient pathway management, workforce productivity, diagnostic services and population health initiatives.

Can organisations can deploy, govern and assure these technologies?

However, as adoption scales, the challenge is no longer simply whether AI can deliver benefits.

The challenge becomes whether organisations can deploy, govern and assure these technologies safely, responsibly and with confidence. 

The publication of specific regulatory guidance for AI-enabled technologies reflects the increasing maturity of the AI ecosystem. As AI moves from isolated innovation projects into mainstream healthcare delivery, governance and assurance arrangements must mature alongside the technology.

Importantly, organisations should be careful not to view governance solely through the lens of regulatory compliance.

Whether a solution falls within medical device regulation or not, there remains a need to understand risks associated with data quality, cybersecurity, information governance, clinical safety, transparency, accountability and supplier assurance. As AI becomes more integrated into decision-making and operational workflows, organisations will increasingly need evidence that these technologies are safe, reliable and delivering the outcomes expected of them. 

This raises a number of important questions for NHS Boards, Audit Committees and senior leaders:

• How are AI-enabled solutions governed and monitored?
• What assurance exists over safety, accuracy and reliability?
• How are information governance and data protection requirements being addressed?
• Who remains accountable for AI-supported outputs and decisions?
• What controls are in place to assure third-party suppliers and technologies?
• How will ongoing performance and emerging risks be monitored?
• How are benefits being measured, evidenced and reported?

Benefits realisation is likely to become an increasingly important consideration. Much of the early discussion around AI has focused on potential gains in productivity and efficiency. As adoption grows, organisations will need robust approaches to demonstrating whether those gains are actually being achieved, whether patient and staff experience is improving, and whether expected benefits are being sustained over time.

Building trust is a key factor

Alongside governance and assurance, there is another factor that will play a critical role in determining the success of NHS AI adoption: trust.

The future of healthcare AI will not be determined solely by the capability of the technology. Clinicians, patients, regulators and organisational leaders all need confidence that AI systems are being deployed transparently, ethically and responsibly. A solution may be technically effective, but if people do not trust how it is being used, adoption will remain fragile. Trust must be designed into AI implementation from the outset, not treated as an afterthought once deployment is complete.

Building that trust also requires investment in AI literacy

As AI becomes more accessible, organisations face a growing responsibility to ensure staff understand both the opportunities and limitations of these technologies. Safe adoption depends not only on the quality of AI systems themselves, but also on users being able to challenge outputs appropriately, recognise limitations, understand risks and know when human judgement must remain paramount. Access to AI does not automatically create AI capability. Understanding remains a foundational component of responsible adoption. 

Recognising this need, MIAA has developed the Simple AI Booklet Series, a collection of accessible resources designed to help NHS staff build confidence and understanding around artificial intelligence. Using plain-English explanations and practical examples, the series explores topics such as AI fundamentals, data quality, bias, governance and assurance, helping organisations create the knowledge foundations needed to support responsible AI adoption.

What defines the next phase of AI adoption?

As the NHS enters the next phase of AI adoption, success will depend on more than choosing the right technology. Organisations will need effective governance, proportionate assurance, informed leadership, strong risk management and a workforce equipped with the knowledge to use AI responsibly.

The conversation is no longer centred on whether AI can support healthcare. The question increasingly becomes how organisations can adopt AI in a way that is safe, trusted, transparent and sustainable.

In that environment, governance is not a barrier to innovation.

It is what gives organisations the confidence to innovate at scale.

Learn more with MIAA's Simple AI Booklet Series:
Simple AI Booklet Series: Helping the NHS Understand Artificial Intelligence