new research fromAppia(Nasdaq: APPN) shows the NHS is the most trusted organization for responsible use of AI, ranking above other public and private sector organizations including banks, retailers and consumer technology companies.
The UK Public Sector AI Adoption Outlook 2026, conducted by Census Wide on behalf of Appian, surveyed 1,000 UK public sector employees and 1,000 UK citizens.
This makes it clear that the NHS stands apart from other public sectors when it comes to public trust in the use of AI.
The NHS is trusted by 63% of the public to use AI responsibly, higher than any other public sector and the only organization with more than 50% net trust.
In contrast, fewer than half of British people trust central government (39 per cent) and local councils (44 per cent) to use AI responsibly.
The NHS also ranks higher than consumer-facing organizations such as banks (55%), retailers (60%) and technology companies (54%).
Overall reliable, but uneven comfort with clinical AI
Confidence in the NHS’s use of AI liability is high, but comfort with specific AI use cases is more mixed.
More than half (56%) of public sector workers say they would be comfortable with AI analyzing NHS scans and diagnostics, compared to just 40% of the public who share this view.
The findings highlight the gap between institutional trust in the NHS and public trust in the use of AI in clinical decision-making.
The NHS is seen as the best place to benefit from the use of AI
A year on from the AI Opportunities Action Plan, in which the government allocated £2 billion to deploy AI research and resources, new research from Appia finds that the NHS is the part of the public sector most likely to benefit from AI investment, with 30 per cent of public sector staff and 29 per cent of the public agreeing.
When asked what benefits they hope to see from integrating AI into public services, citizens top the list with faster services and lower waiting times (35 percent), improved safety (26 percent), and easier-to-use digital services (26 percent).
Communication with patients must be prioritized
The report’s findings highlight the need for NHS organizations to combine clinical and non-clinical AI deployment with clear explanations, human oversight and demonstrable safeguards, especially as public understanding of how AI is used in the public sector remains low.
In fact, 75 percent of the public cannot name a single way their government is currently using AI, suggesting that there is a significant communication and transparency opportunity for governments to publish accessible explanations of where and how AI is being used.
Is the NHS ready?
Despite high public trust, levels of organizational readiness within the NHS are relatively low, with only 6% of healthcare workers saying their organization is ‘fully capable’ of leveraging AI, indicating a significant gap in implementation.
A majority of respondents said process preparation and “fundamental process fixes” should take priority over AI implementation, with both public sector employees and the public agreeing that this should be done before introducing new AI technology (55 percent and 56 percent respectively).
“The NHS is widely recognized by both public sector workers and the public as the area of the public service most trusted and most likely to benefit from AI,” said Peter Cope, UK public sector industry leader at Appian.
“To close the gap in healthcare delivery, healthcare leaders must avoid the temptation of the ‘shiny new toy’ and instead look to improve the patient journey with AI embedded in core processes, giving it purpose, guardrails and goals, and making it effective, safe and measurable.”
The report confirms that not all public sector organizations are applying AI to improve established workflows.
Some companies are adopting it as a bolt-on experiment (23 percent) or as a standalone work tool (22 percent) with little or no integration with existing core processes.
This suggests that 45% of government AI applications are not implemented as part of integrated process improvement, creating potential challenges in terms of delivery and impact.
This lack of integration into existing processes is one reason why MIT found in its 2025 report that 95% of AI investments did not yield a return.
Cope concludes: “AI works when tasked with a controlled process that moves beyond experimentation to an integrated and measurable program.
“For healthcare leaders, the advice is clear: Give AI the right data, guardrails, and purpose, and your organization will deliver results that are secure, auditable, and accessible to the public.”
“The key is to start with the process and incorporate AI into that process.”
download Outlook for UK public sector AI adoption in 2026 For detailed survey results.
