Data governance and community accountability in the use of AI in public health agencies

The gap

Evidence profile

Sourced from the recommendations and future work and stated research gap of the source papers, classified as general, spanning 3 journals. Those papers have been cited 1 times in total.

Supporting evidence — 3 representative gaps

• Perspectives on healthcare artificial intelligence policy from health equity professionals: findings from an interview study (2026) · Frontiers in Digital Health · doi

  Theme 1: Developing health AI policy 1. Increase community diversity and representation in datasets, add “warning label” to datasets which lack diversity for health equity starts with data 2. Learn about how social contexts and biases are embedded in health data 3. Identify power asymmetries in data collection and build structures for community access to health data for AI 4. Encourage collection and inclusion of data that gives a fuller picture of health and life history, such as data from caregivers Theme 2: Health AI policy for health 1. Longer term commitments from funders for AI and health disparities research, also include small research institutes; equity must include multiple encourage academic and community research partnerships to investigate health equity in AI institutions and strategies 2. Encourage federal agencies to investigate how health AI tools may violate rights 3. Include robust evidence on AI benefit and equity as part of regulating health AI, such as clinical trials as part of FDA AI regulation 4. Encourage accreditors such as Joint Commissions to include health equity in assessment of health AI Theme 3: Considering economic issues 1. Use AI tools to identify high-needs patients without expecting organizations to “do more with less” is key for developing health AI policy 2. Pitch equitable AI as a marketing strategy that advances health equity 3. Invest in under-resourced healthcare facilities to increase capacity to use and maintain AI solutions 4. Consider how integration of health AI tools can have workforce implications getting health data can be an “extractive” process, and that measures should be taken to ensure that there is mutual benefit: tools. The the data collection “communities who want the data, who are generating the data, and that either they collect about themselves or that others collect about them, you have large institutions that, you know, deploy large organizations benefit from the use of that data, but again they’re larger organizations that have a lot more power in the communities too, so, for one, recognizing that the communities have access to the data that’s being collected about them. Access such that they don’t need to overcome paywalls or overcome significant technical boundaries that might exist around them accessing data collected about them … Do the pipelines exist for communities to benefit from that data in a way that empowers and enriches? In [a way] that helps communities address, you know, the needs that they have right then and now … organizations that collect data or extract data from and about communities. What are they responsible for? Who’s holding them

  generalrecommendations
  Keywords: health equity communities them encourage include tools bene organizations theme policy community collection access collect
• A Mixed-Methods Study of Policymakers’ Adoption of AI to Support Use of Research Evidence: Implications for Artificial Intelligence in Prevention Policy (2026) · Prevention Science · cited 1× · doi

  Building on the promising early adoption of the Results First AI Assistant, several avenues for future work are recom- mended. First, expanding the tool’s capabilities to include additional data sources and intervention domains could broaden its utility for policymakers. For example, integrat- ing databases related to education, criminal justice, and pub- lic health could provide a more comprehensive resource for legislators addressing complex, cross-sector policy issues. Second, long-term evaluations are needed to assess the AI assistant’s impact on policy decision-making processes and outcomes. Future studies should examine whether access to the tool leads to increased adoption of evidence-based interventions, improved allocation of resources, and measur- able benefits for target populations. Such evaluations would provide critical evidence on the tool’s effectiveness and inform ongoing refinement. Third, further exploration of the ethical implications of AI use in policymaking is war- ranted. While the AIRE Protocol addresses key concerns related to transparency and bias, continued engagement with ethicists, community stakeholders, and policymakers is essential to navigate emerging challenges. Issues such as algorithmic accountability, equity in access to AI tools, and the potential for overreliance on automated recommenda- tions should remain central to future development efforts. Fourth, efforts to scale the tool beyond state legislatures to include federal policymakers, local governments, and inter- national policy bodies could enhance its reach and impact. Collaborations with organizations such as the National Conference of State Legislatures and the Council of State Governments have demonstrated the value of partnerships in facilitating tool dissemination and adoption. Finally, as AI technology evolves, continuous improvement processes must be maintained to ensure that the AI assistant remains 1264 Prevention Science (2026) 27:1253–1267 aligned with best practices in both prevention science and AI development. Regular updates to the tool’s evidence base, user interface, and algorithmic components will be essential to sustain its relevance and effectiveness in dynamic policy- making environments.

  generalfuture work
  Keywords: tool policy adoption assistant future policymakers evidence state first include related provide issues evaluations impact
• Public Health Responsible AI Capability (PH-RAIC) Framework: A Conceptual Model for Integrating AI into Public Health Agencies (2026) · Healthcare · doi

  The gap in data governance and community accountability in the use of AI in public health agencies. The lack of a conceptual framework for responsible AI use in public health agencies, given the fragmented health data collection and stewardship. The need for a framework that adapts principles of transparency, accountability, fairness, ethics, and safety to institutional realities faced by public health agencies.

  generalstated research gapevidence 5/5
  Keywords: gap data governance community accountability use public health

Questions about this gap