A new paper explores how legal liability for AI in healthcare can influence both algorithm design and physician usage. The study suggests that a liability rule in the United States, which holds providers responsible for errors stemming from disparate AI accuracy, may lead physicians to reduce their reliance on AI, particularly for disadvantaged patient groups. This, in turn, can alter the AI firm's design incentives, potentially leading to less investment in reducing disparity or even a shift towards equal-accuracy designs, which could inadvertently harm both patient groups. AI
IMPACT Examines how legal frameworks can shape the development and deployment of AI in critical sectors like healthcare.
RANK_REASON Academic paper analyzing the impact of policy on AI design and use. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
