1. Providers are using unauthorized AI to streamline workflows
Among users who admitted to using unauthorized AI tools, 45% said the main reason was to speed up their workflow. A further 24% said the main reason was because it had better functionality than currently approved tools, supporting the need for clinician input in corporate tool selection.
2. Providers are more likely to experiment with unapproved AI tools
One of the main areas where providers and administrators differed was in tool experimentation. Among users who have used unauthorized tools, 26% of providers did so for curiosity or experimental purposes, compared to 10% of administrators. Without established corporate standards, clinicians will experiment with new tools that fit their workflows.
3. AI tools are being used by providers across all lengths of care
Many organizations are working to use unapproved AI tools, primarily targeting trainees and young clinicians. However, among providers with less than five years of experience, only 14% have used an unapproved AI tool, and 16% said they have never used an AI tool. Of those with 5+ years of experience who have used an unapproved AI tool, 42% used it to speed up their workflow, 29% for curiosity and experimentation, and 23% because it provided better functionality than an approved solution.
A company-wide AI policy is key
It is important for leaders to clearly establish and communicate AI policies throughout the enterprise. It is also important that providers participate in the selection of AI tools to ensure evidence-based solutions are considered. Without input, providers may revert to using their preferred tools behind the scenes.
Learn more about our findings in our free white paper, Shadow AI: The Hidden Risks of Healthcare.
