Human-AI Collaboration in Decision-Making: Opportunities and Risks in Healthcare

Abstract

The incorporation of artificial intelligence (AI) into clinical workflows heralds a paradigm shift in human-AI collaboration in health care. AI solutions provide important decision-support advantages ranging From diagnostics to personalized treatment planning. That said, these developments also bring with them concerns of trust, accountability and ethical deployment.We explore these domains by considering the opportunities and the dangers of human–artificial intelligence decisions for healthcare, which we contextualise with an emphasis on diagnostic performance, ethics and clinical implementation.We performed a literature review and a perception-based survey of 110 healthcare professionals. The results were synthesized to provide a balanced perspective on how AI enhances the abilities of the human expert, while simultaneously creating some vulnerabilities.Results show that AI improves speed and precision in diagnosis, but increases data bias, decreases human control, and creates ambiguity regarding liabilityOpen human-AI collaboration can provide considerable value at an institutional scale, but must be underpinned by rigorous ethical standards, transparent processes and collaborative processes across disciplines in the field of health care systems.It comes down to human-AI collaboration in healthcare decision-making