Harnessing public perception to drive responsible AI in healthcare
Nell Thornton, improvement fellow at The Health Foundation, looks at the importance of public and staff perceptions of innovations like AI.
As healthcare continues to embrace technology, understanding public and staff perceptions of innovations like AI is not just a nice-to-have – it is essential. At The Health Foundation, we have made it a priority to delve deeply into these perceptions to ensure that any technological advancements made within the NHS are inclusive, widely supported, and beneficial to everyone.
Over the past two years, we have conducted extensive research, surveying both the public and NHS staff to explore attitudes towards technology and data use in healthcare. Our findings reveal that the public perception of AI in healthcare is finely balanced. A little over half (54%) of respondents support the use of AI in patient care and 61% back its use for administrative purposes. These numbers reflect a cautious optimism, with people more open to AI handling non-clinical tasks than taking a more direct role in patient care. This suggests a latent receptiveness that, if nurtured, could pave the way for AI to revolutionise certain aspects of healthcare delivery.
However, there are important caveats to this support. Concerns about the human element of care being diminished were particularly prevalent. These findings underscore the need for AI implementations to be designed in ways that preserve and enhance the relational aspects of care.
There are also concerns around decision-making and the accuracy of AI outputs. The public and staff alike are wary that AI-generated decisions could go unquestioned. Ensuring that AI is subject to human oversight is key to maintaining public trust, as our research showed that support for AI dramatically drops when there is no human in the loop.
Staff perceptions
NHS staff, overall, are more supportive of AI than the public. Despite sharing many of the same concerns, they are generally more comfortable with its integration into healthcare processes. This likely reflects their experience and familiarity with healthcare systems, and a greater understanding of how AI could be beneficial in administrative tasks, potentially alleviating some of their daily burdens.
The role of language
Our research also extended to exploring public perceptions of newer healthcare models, such as virtual wards. Here, we found that terminology plays a critical role. While the term virtual wards garnered a moderate level of support (45%), when we reframed the question to ask whether people would be happy to monitor their health at home using devices like blood pressure monitors, support jumped to 76%. This disparity highlights the need to consider how we communicate new models of care. Inequalities in access
Another significant area of concern is the equity of access to new technologies. Our research consistently shows that those in socio-economic groups are less supportive of technology in healthcare. In the case of virtual wards, for example, people from lower-income backgrounds often cited concerns that their homes would not be suitable for such a model of care. This reflects a broader issue of health inequity that extends to digital literacy, access to suitable housing, and the resources required to benefit from healthcare technology.
It is crucial that as the NHS rolls out technologies like AI and virtual wards, it works closely with social care and housing teams to ensure that these innovations are accessible to everyone.
Responsible AI
While there are many promising initiatives, they are currently fragmented, leading to a scenario where the whole is less than the sum of its parts. We need to see greater collaboration and coordination to ensure that the AI tools being developed and deployed are robust, fair, and, above all, responsible.
Responsible AI is about more than just minimising harm; it is about ensuring that the benefits are equitably distributed across society. This includes addressing the biases in AI systems, ensuring transparency in how decisions are made, and involving patients and the public in the co-design of these technologies. Only then can we harness the full potential of AI to improve healthcare for everyone.
Nell Thornton will be discussing the future of healthcare technology and its role in shaping a more inclusive, efficient NHS at Rewired 2025 on 18-19 March 2025.
