Hospitals favour hybrid AI over fully autonomous systems, survey finds

Summary

A new Carta Healthcare survey of hospital and health-system executives finds strong scepticism toward fully autonomous AI in clinical environments, with respondents overwhelmingly preferring models that operate under clinician oversight. Only 12.5% of organisations said autonomous AI had delivered the most value in their work so far, while 75% reported relying on human validation to ensure AI outputs are clinically relevant and trustworthy. Top concerns about autonomous systems included misinterpretation of data and "black box" decision-making misaligned with established risk scores, both seen as direct threats to patient safety. The report frames "hybrid intelligence"—AI plus expert oversight—as the winning pattern for near-term care settings, echoing a broader industry shift away from hypey automation toward augmentation. For AI vendors selling into healthcare, the message is blunt: products that ignore workflow integration and human-in-the-loop checks are unlikely to gain traction with risk-averse clinical buyers.