New research from Censuswide (n=2,000, April 2026) reveals that healthcare providers face a critical trust gap — not because patients reject AI, but because most don’t know it’s already happening.
AI is already transforming how UK patients are diagnosed, triaged and treated. It analyses your X-ray before a radiologist reads it, it prioritises who the GP sees next, and it flags anomalies in your blood test results before the doctor picks up the phone. And yet, according to our new research, the majority of UK adults either don’t believe this is happening or simply aren’t sure.
That blind spot is the single biggest reputational risk facing healthcare brands today.
Do UK patients know AI is involved in their care?
More than half of UK adults (55%) either don’t believe that AI is currently being used to deliver healthcare in the UK — including within the NHS — or say they’re not sure. Just 45% believe it is.
Among those who do recognise AI’s presence, administrative tasks (52%) and medical image analysis such as X-rays and MRIs (45%) are seen as the most likely applications. Fewer connect AI to diagnostic support (39%) or treatment planning (30%), even though these are among the most consequential uses already in deployment.
The awareness gap is significant in itself. But the real story is what happens to trust when that gap closes — and how patients find out.
What do patients want from healthcare providers using AI?
84% of UK adults believe healthcare providers should tell patients when AI has been used in their diagnosis or treatment . Of those, nearly two thirds (62%) say providers should always disclose this, not just in certain circumstances.
Only 7% say providers should never disclose AI involvement.
When asked specifically which situations warrant disclosure, patients don’t draw a distinction between serious and routine care. Disclosure expectations are high whether AI plays a major role in decision-making (54%), in the context of serious or life-threatening conditions (43%), or even in routine diagnoses (41%).
The message to healthcare brands is clear: patients aren’t asking for AI to be removed from their care. They’re asking to be told.
Would patients opt out of AI in their healthcare?
Given the choice, 43% of UK adults say they would opt out of AI being used in their care. Less than a quarter (23%) say they would not opt out, and a further 31% remain undecided.
The opt-out preference is broadly consistent across genders, but younger patients — particularly Gen Z (18–28) — are more likely than average to say they would opt out (51%), [IA1.1]which challenges assumptions that digital-native generations are uniformly positive about health technology.
How does AI involvement affect patient confidence and provider trust?
This is where the data becomes most instructive for marketing and communications teams.
When patients are told that AI was involved in their care, 38% say they would feel less confident in the care they were receiving — significantly more than the 24% who say it would make them more confident . A sharp gender divide emerges here: men are significantly more positive (31% say more confident) than women (17%), while women are more likely to feel less confident in AI-assisted care.
The more pressing question, however, is what happens to trust when patients find out rather than being told upfront. If a patient discovers that AI played a role in their diagnosis without prior knowledge, nearly half (48%) say it would decrease their trust in their healthcare provider . Just 19% say it would increase trust .
But here’s the critical insight for communications strategies: if patients are informed in advance that AI will be used, the trust picture improves significantly. Only 34% say their trust would decrease when told proactively — a 14-percentage-point improvement over the retrospective discovery scenario. And the proportion saying it would increase trust rises from 19% to 25% .
Transparency has become a measurable trust asset.
Would patients choose an AI-free healthcare provider?
Given a direct choice, 37% of UK adults say they would prefer a healthcare provider that does not use AI at all. That edges out the 32% who would choose a provider that uses AI and clearly communicates this, and dwarfs the 17% who would accept AI use without clear communication.
Women are more likely than men to prefer an AI-free provider (40% vs. 33%), reflecting the broader gender gap in AI confidence seen throughout the data.
For healthcare brands actively investing in AI capabilities, the implication is stark: the competitive advantage of AI adoption can only be unlocked through clear, proactive communication. Without it, the technology that is meant to differentiate a brand risks driving patients towards competitors who simply don’t use it.
Will patients share health data to improve AI-assisted care?
The one area where sentiment turns measurably more positive is data sharing. 61% of UK adults say they would be willing to share more personal health data if it improved the accuracy of AI-assisted diagnosis and treatment . Only 25% say no outright.
However, willingness is largely conditional: of those who say yes, the majority (41% of all respondents) say their willingness depends on how the data is used. Just 20% say they would share without conditions.
The barriers are significant. Data privacy and security concerns top the list (46%), closely followed — and this is notable — by fear of losing human involvement in their care (44%). Worries about AI accuracy (40%) and misuse of data by insurers or third parties (37%) also rank highly.
That fear of losing the human element may be the most important signal in the entire dataset. Even among the 61% willing to engage with AI-driven healthcare, the expectation is clearly that AI augments human care rather than replaces it. Healthcare brands that position AI as a tool that frees up clinician time — rather than one that stands in for it — are likely to resonate far more strongly.
What does this mean for healthcare marketers?
Our research points to a clear strategic imperative: the question for healthcare brands is no longer whether to use AI, but how to communicate it.
Patients are not categorically opposed to AI in healthcare. They believe it can reduce waiting times (32%), accelerate diagnosis (31%) and detect disease earlier (28%). But more than one in four (27%) currently see no benefit from AI in healthcare at all — a figure that should concern any brand yet to invest in patient education around its AI capabilities.
The brands that will lead are those that close the awareness gap proactively, build disclosure into the patient experience as standard practice, and centre the human clinician — not the algorithm — in how AI-assisted care is communicated. The data shows that transparency is not a risk to be managed. It is the most effective trust-building tool available.
Research conducted by Censuswide among 2,000 nationally representative UK adults aged 18+, 24–27 April 2026.
¹ Combined result of ‘Yes – always’ and ‘Yes – but only in certain situations’.
² Combined result of ‘Slightly less confident’ and ‘Much less confident’.
³ Combined result of ‘Slightly more confident’ and ‘Much more confident’.
⁴ Combined result of ‘Decrease trust slightly’ and ‘Decrease trust significantly’.
⁵ Combined result of ‘Increase trust slightly’ and ‘Increase trust significantly’.
⁶ Combined result of ‘Decrease trust slightly’ and ‘Decrease trust significantly’.
⁷ Combined result of ‘Increase trust slightly’ and ‘Increase trust significantly’.
⁸ Combined result of ‘Yes, definitely’ and ‘Yes, depending on how the data is used’.
