November 5 — About 94% of Chinese hospitals now use at least one AI product, and AI penetration exceeds 50% in each of the three core scenarios — medical imaging, smart outpatient services and clinical decision support — according to a Guandian report, marking the shift from pilot wards to routine operation.

AI in the hospital: imaging workstation with an assistant-diagnosis data network
AI in the hospital: imaging workstation with an assistant-diagnosis data network

The pivotal change came on the payment side. In late 2024, China’s National Healthcare Security Administration listed AI-assisted diagnosis as a chargeable item in its pricing guidelines, and from April 2025 the service was rolled out under national basic medical insurance. Once the fee catalog connected to insurance settlement, AI diagnosis stopped being a hospital’s research hobby and became a billable clinical service — changing the financial logic of every purchase decision.

Key data points

Hospital coverage: 94% use at least one AI product

Scenario penetration: above 50% in imaging, smart outpatient and decision support

Surgical robots: penetration rose from 3.7% (2020) to 11.2% (2025); 68% in minimally invasive surgery

Market outlook: China surgical-robot market to exceed 28.7 billion yuan by 2030 (Frost & Sullivan)

Operating rooms show the same climb. Intuitive Surgical performed roughly 3.2 million procedures worldwide in 2025, while China’s surgical-robot penetration rose from 3.7% in 2020 to 11.2% in 2025 (IIM data), with 68% of use in minimally invasive surgery — roughly a tripling in five years across both installed base and procedure volume.

Forecasts follow. Frost & Sullivan expects China’s surgical-robot market to top 28.7 billion yuan by 2030, as domestic vendors win approvals across laparoscopic, orthopedic and natural-orifice segments and loosen imported equipment’s grip on top-tier hospitals.

Coverage, however, is not value. Hospital AI still clusters around image reading; deep applications spanning treatment, medication and follow-up remain rare. The layered penetration reflects three thresholds — data governance, cross-department workflow and liability allocation — and the number of departments that clear them defines AI’s true depth inside each hospital.

Insurance reimbursement offers a replicable commercial template: per-item pricing settled with the diagnosis. Whether it extends to more clinical AI hinges on how fast efficacy evidence accumulates — health-economic evaluation becomes the next battleground, and only products that prove shorter stays or fewer complications will hold pricing power in the next round.

From 94% coverage to 50%-plus scenario penetration, Chinese medical AI has finished the first half of its journey; the second half is converting usage into measurable clinical and economic returns.