1. Determine barriers and opportunities for “bottoms-up” AI adoption in clinical and administrative settings.

2. Identify at least three concrete ways specialty practices are using AI today.

3. Apply best practices for integrating AI-enabled platforms into existing practice operations to reduce administrative burden and improve patient care. 

Artificial intelligence is often framed as a “top-down” innovation, deployed by payers, regulators, and hospital systems. This session explores how AI can be leveraged in a  “bottoms-up” way to solve the administrative challenges that burden physicians, staff, and ultimately, patients.

SamaCare will share survey findings on how practices are incorporating AI into daily operations. Attendees will walk away with practical strategies, data-driven insights, and an understand of the role of AI at the practice level.

We conducted a structured survey of specialty practices within SamaCare’s network. This network spans nearly all 50 states, representing a diverse cross-section of practice sizes, ownership, and geographies.

Respondents included practice administrators and managers, prior authorization specialists, and senior executives.

Responses were collected using a secure online survey platform. Participants were compensated for their time.

Our survey revealed a surprising breadth of experimentation with AI among practices. Practices are exploring creative use cases, often leveraging open-source tools, alongside enterprise platforms.

Key insights:

• Unexpected use cases: Several practices reported using AI for payer appeal letters, summarize prior authorization requirements, or generate patient-facing educational materials. Respondents estimated these tasks took 50–70% less staff time with AI, freeing teams for higher-value tasks.
  
  

• Wide range of maturity: Adoption ranged from no AI exposure to sophisticated tactics using open-source and enterprise tools.
  
  

These findings suggest that for practices, AI can save time, reduce workload, and accelerate patient access.