The trend: Many physicians and hospital staff encounter or use “shadow AI” tools (those that haven’t gone through the proper approval processes) at their organizations, according to a December 2025 Wolters Kluwer Health survey of 518 healthcare providers and administrators.
- 40% of respondents saw an unauthorized AI tool in their organization, but did not use it.
- 17% admitted they have used unapproved AI tools themselves.
- About 1 in 10 said they’ve even used an unauthorized AI tool for direct patient care.
Why it matters: AI use is rapidly expanding across hospitals and health systems, with physicians using both general-purpose tools and healthcare-vetted solutions.
However, healthcare organizations typically vet and select AI products methodically—reports of unauthorized usage expose gaps in some hospitals’ AI strategies. About 80% of healthcare C-suite leaders say they’re either beginning to implement AI or cautiously evaluating tools, while just 10% report aggressively pursuing AI, per an August 2025 Sage Growth Partners survey.
As executives weigh AI investments, clinicians and staff may be tempted to use readily available tools without vetting them.
- 50% of respondents who used unapproved AI products in Wolters Kluwer’s study said they did so for a faster workflow.
- About one-third noted either a lack of approved tools or the approved tools lacking the desired functionality.
Implications for healthcare provider organizations: Leading AI companies like OpenAI and Anthropic are developing tools to improve doctors’ workflows, setting up competition with established, specialized AI players. As AI solutions flood healthcare, hospitals and health systems must put AI governance in place, setting clear guidelines for responsible, general-purpose AI use in patient care, while also involving clinicians in testing tools before investing.