AI that actually helps your practice: 7 tips for administrators

In 2026, it feels like artificial intelligence (AI) is everywhere. It’s embedded in ambient documentation tools, coding platforms and contact centers, and often turned on by vendors long before physicians and practice leaders have a clear strategy for what they want it to do.

According to the "Physicians on AI: 2025 Trends and Insights" survey by HealthLink Dimensions, although 40% of physicians are excited about AI’s potential, 41% still don’t use any AI tools at all. That split shows up in day-to-day conversations with practice leaders who see genuine value in automation but worry about safety, bias, liability and cost.

• Related content: “The new malpractice frontier: Who’s liable when AI gets it wrong?”

In a Medical Group Management Association (MGMA) Stat poll of 177 applicable respondents cited in “Automatic for the people: AI moves for medical practices to boost the front office and access”, 31% of practice leaders said scheduling is the top area they want to automate with AI, followed by calls (27%), registration and eligibility (23%) and prior authorization (16%).

For independent practices and health systems alike, the question is no longer whether AI will show up, it’s how to decide where to use it first, what to say no to and how to build guardrails that protect patients, clinicians and the business.

This slideshow outlines seven concrete moves physicians, practice leaders and administrators can make to put AI to work in manageable, defensible ways.