Between Fear and Adoption: The AI Paradox in Medico-Legal Practice Among National Insurance Physicians

Area of Science:

• Medical Informatics
• Health Services Research
• Legal Medicine

Background:

• Artificial intelligence (AI) is increasingly adopted in clinical settings.
• Its use in insurance medicine, particularly for disability assessments, is underexplored.
• Physicians in this field face unique challenges at the intersection of clinical, administrative, and legal demands.

Purpose of the Study:

• To evaluate AI adoption patterns, trust levels, and perceived utility among physicians performing disability assessments.
• To investigate the disparity between AI's widespread use and limited trust in medico-legal contexts.

Main Methods:

• A cross-sectional survey was administered to 428 physicians at the Israeli National Insurance Institute.
• The survey assessed AI usage frequency, trust in AI decisions (5-point scale), perceived risks, and preferred applications.
• A high response rate of 85.6% was achieved.

Main Results:

• AI adoption was prevalent, with 76% using it regularly or occasionally and 27% daily.
• Trust in AI decisions was low (mean 2.7/5), though daily users reported higher trust (OR 1.81).
• Key drivers for adoption included administrative burden, with preferred uses like record summarization (85%) and inconsistency detection (45%).

Conclusions:

• Physicians differentiate between instrumental trust for efficiency and epistemic trust for legal accountability.
• AI integration in insurance medicine should prioritize administrative support while safeguarding human interpretative authority.
• Future AI development should address both relational and technical concerns raised by physicians.