When the American Medical Association first asked physicians about AI in 2023, 38% said they were using it professionally.

By March 2026, that number was 81%.

That's not incremental growth. That's a doubling in three years. And the implications for small and solo practices are specific enough to be worth talking through directly.

What Doctors Are Actually Using AI For

The AMA surveyed nearly 1,700 physicians across specialties and practice settings. The most common uses:

• Staying current on medical research - AI as a reading and synthesis tool
• Creating discharge instructions - drafting written patient communications
• Documenting visits - the ambient scribe use case that's been growing for two years

Documentation leads the adoption curve for a reason: it's the most painful part of running a small practice that doesn't directly help patients. The average primary care physician spends roughly two hours each evening finishing notes. Tools that compress that to 20 minutes are not a nice-to-have.

The Confidence Shift

What's notable in the AMA data isn't just the adoption rate - it's that physicians are now more comfortable, not less, after using these tools. In 2023, many doctors were skeptical of AI accuracy and worried about patient privacy. Three years of daily use has shifted that calculus.

AMA CEO Dr. John Whyte framed it this way: "AI has quickly become part of everyday medical practice. Physicians see real promise in its ability to support clinical decisions and cut down on administrative burden."

The key word in that sentence is "support." The AMA's position is that AI should augment physicians, not replace clinical judgment. That's not just a political hedge - it's the actual design philosophy of the tools that are working in practice.

What This Means for Small and Solo Practices

Large health systems have IT departments. They have vendor relationships and implementation budgets. For a solo family practice or a two-physician specialty group, the path to AI adoption is harder to navigate.

Three areas where the ROI is clearest for small practices:

1. Ambient documentation. Tools like Nabla, Abridge, and Suki listen to your patient conversations and generate draft notes automatically. You review and approve. This is the highest-impact AI application for small practices right now - not because it's glamorous, but because it buys back time you currently spend after hours.

2. Prior authorization assistance. This is the hidden time sink in small practices. AI tools are now available that pre-fill authorization requests, track denials, and draft appeal letters. For practices doing 20+ authorizations a week, this can save 5-8 hours of staff time.

3. Patient communication drafts. After-visit summaries, prescription instructions, referral follow-ups - AI can draft these in seconds. Your staff reviews and sends. The patients get clearer information. You get fewer callback calls.

The Honest Concern: Cost and Fit

The AMA survey also flagged that cost remains a notable barrier for smaller practices considering integrated AI solutions. A full ambient transcription system with EHR integration can run $300-600 per month per provider. For a solo practice with thin margins, that's not trivial.

The decision framework worth applying: What is one hour of your time worth? If you are spending 10 extra hours per week on documentation and an AI tool could recover 7 of those hours, the math usually works even at $500/month. But you have to do the math for your specific volume and billing rate.

What the AMA Is Pushing For

Beyond tracking adoption, the AMA is now advocating for payment policies that improve AI affordability for small practices. They want payers to recognize AI-assisted care in billing structures, and they've pushed for insurance reimbursement models that don't penalize smaller practices for adopting new tools.

That policy work moves slowly. But it signals that the major physician organizations are treating AI as infrastructure for small practices, not just a premium feature for large health systems.

The Bottom Line

81% of physicians using AI doesn't mean 81% of small practices are running sophisticated systems. A lot of that adoption is ChatGPT for research questions or basic drafting. But the directional signal is clear: the floor has risen.

If you're running a small practice and haven't yet tried an AI documentation tool, you're now in the minority of your peers. That doesn't mean you need to buy the most expensive system on Monday. It means a pilot - even a free-tier test of one tool for 30 days - is overdue.

The research is done. The adoption curve has turned. The only question left is how long you want to wait.

Sources: AMA 2026 Physician Survey on Augmented Intelligence | AMA Center for Digital Health and AI