Integrating Artificial Intelligence (AI) into healthcare requires a social licence to operate. The need for extensive health data to train AI, combined with shifting patient-clinician relationships, sparks critical debates regarding trust and benefits from AI innovations.
While participants view the “AI turn” as inevitable, they express anxiety over a lack of agency about how this “AI turn” will progress. They expect to be included in setting the direction of this transformation. Claims of “push-button” efficiency are met with skepticism. Consumers fear this may lead to the oversimplification of complex health needs, further reduction in human resources and the erosion of human care. Participants recognize that AI is in its infancy and requires communal data contributions. However, they view this as reciprocal: powerful institutions cannot expect data access without giving the community a say in shaping the rules. Willingness to contribute data is often driven by expected future benefits, emphasizing the need for a Data Solidarity framework that prioritizes public value.
What does this mean for policy and practice?
1. Anchor AI centric health care innovation in the Quintuple Aim for better health, better care, lower costs, clinician well-being, and equity. AI should reduce administrative burden and support care given by medical experts. Efficiency gains are to be reinvested into human relationships and improved health care experience. We call this the ‘AI-in-the-Loop’ principle.
2. Make explicit what data are being used, who uses the data, for what purpose and how. This serves to demonstrate that AI is feeding back into better care and minimizes exaggerated claims and unrealistic fears. Provide concrete and accessible examples of tangible benefits and acknowledge that AI is a tool to manage complexity, not a magic solution to oversimplify healthcare. We call this the ‘Observability’ principle.
3. Treat data as a public good. Establish partnerships where commercial engagements align with the core values encapsulated in the Quintuple Aim. Foster acceptance of individual costs (i.e., sharing data) among health consumers due to collective benefits that will be generated (i.e., the need for good care). We call this the ‘Data solidarity’ principle.
4. Move beyond consultation to active participation. Involve the community directly in defining the “rules of the road,” determining which AI innovations to prioritize, and deciding how the benefits of AI are distributed. We call this the ‘Govern through Community Co-Design’ principle.