UK Study Reveals Conditional Support for Health Data Sharing in AI

Public support for sharing health data for artificial intelligence (AI) research is contingent on clear benefits, robust safeguards, and meaningful consent, according to a comprehensive study conducted by researchers from the National Institute for Health Research (NIHR). This study, published in BMJ Digital Health & AI, is based on in-depth discussions with members of the U.K. public and reflects growing concerns and considerations about health data sharing.

The research highlights the complex landscape of public opinions surrounding the use of health data in AI applications. Lead author Rachel Kuo, a NIHR Doctoral Research Fellow, emphasized the increasing integration of AI into healthcare and the necessity for access to extensive patient data for its development. “Our aim was to understand how people think about sharing their data in the context of AI,” Kuo stated, noting the public’s concerns regarding confidentiality and security.

Key Findings from Focus Groups

The researchers conducted eight online focus groups involving a total of 41 participants, representing diverse ages, ethnicities, health backgrounds, and socioeconomic statuses across the U.K. Participants engaged in discussions about various health data-sharing scenarios, including those involving academic research, large databases, and commercial projects. From these discussions, three prominent themes emerged.

First, participants exhibited cautious and conditional support for sharing health data. They recognized the importance of data anonymization but also expressed skepticism about its effectiveness, particularly concerning individuals with rare conditions. Many participants accepted that some risk is inherent in data sharing but emphasized the need for transparency regarding data protection measures and protocols in the event of breaches. Trust varied significantly depending on the data user. While universities and the National Health Service (NHS) generally inspired confidence, skepticism rose with the involvement of commercial organizations unless a clear public benefit was established.

The second theme highlighted individual decision-making processes regarding data sharing. Participants weighed their perceived risks against the potential benefits, including improved healthcare outcomes, expedited diagnoses, and contributions to future medical advancements. Many expressed a strong sense of responsibility toward the well-being of others, particularly those with chronic conditions, which motivated their willingness to share data.

Lastly, the importance of informed consent emerged as a cornerstone of public trust. Participants indicated a desire for clear, relevant, and accessible information related to specific studies when consenting to data sharing. They advocated against consent requests made during emotionally charged clinical moments, suggesting tailored approaches, cooling-off periods, and the ability to withdraw consent later.

Community Involvement and Trust

A notable aspect of the study was its collaborative approach, involving patient and public involvement (PPI) contributors who played a pivotal role in shaping the research questions, conducting focus group interviews, and analyzing the data. This methodology ensured that the study addressed genuine public concerns, enhancing participant confidence and eliciting authentic opinions.

Rosie Hill, a PPI co-producer, remarked on the significance of the research in capturing public sentiment. “This is very important work that speaks directly with the public to understand the views that really matter,” she stated. Another PPI co-producer, Judi Smith, highlighted the complex dynamics at play, noting that while some participants had reservations about sharing data, particularly with commercial entities, others were open to sharing if it could accelerate treatment for their conditions.

Kuo concluded that as healthcare systems increasingly rely on large-scale data to develop AI technologies, public trust cannot be assumed. “Our research shows that people are willing to support data sharing, but only under clear conditions,” she stated. These conditions include transparency about data use, strong governance, meaningful consent, and demonstrable public benefits. Recognizing these expectations will be essential for ensuring ethical and sustainable innovation in healthcare.

The study, titled “Public perceptions of health data sharing for artificial intelligence research: a qualitative focus group study in the UK,” is set to be published in BMJ Digital Health & AI in 2026 and serves as an important contribution to the dialogue around health data sharing in the context of advancing technology.