Towards responsible surveillance in preventive health data-AI research

Area of Science:

• Health Informatics
• Bioethics
• Artificial Intelligence

Background:

• Artificial intelligence (AI) integration in health data research offers potential for precision medicine and managing chronic conditions.
• However, health AI also significantly advances surveillance, raising substantial ethical and social concerns.
• Health surveillance, particularly concerning data-AI research and innovation, remains an understudied area.

Purpose of the Study:

• To provide a conceptual analysis of health data-AI surveillance.
• To examine the evolution and amplification of surveillance practices through AI in healthcare.
• To assess the implications and challenges of health data-AI surveillance and propose responsible governance strategies.

Main Methods:

• Conceptual analysis of health data-AI surveillance.
• Case study utilizing the Hypermarker research project.
• Tracing the historical evolution of surveillance in medicine, public health, and digital health technologies.
• Analyzing AI's amplification of existing surveillance practices and their implications.

Main Results:

• Health data-AI surveillance presents implications such as pervasiveness, hypercollection, function creep, hypervisibility, profiling, and the formation of surveillant assemblages.
• The Hypermarker project implemented safeguards and measures to address these challenges.
• Key challenges identified include strengthening trustworthiness (fairness, equity), ensuring accountability (transparency), and fostering public control and oversight.

Conclusions:

• Advancing responsible governance in health data-AI research is essential.
• Recommendations include implementing community advisory panels, independent review boards, data-AI justice frameworks, transparency dashboards, and open oversight cycles.
• Strengthening trustworthiness, accountability, and public control are critical for ethical health data-AI surveillance.