The stats might shock some CTOs and compliance officers.
According to the American Medical Association’s most recent Physician AI Sentiment Report, more than 80% of physicians now report using AI in some professional capacity. That represents a dramatic increase over prior years, and it may signal that practices and health systems need to catch up.
Today, adoption spans documentation support, clinical research, coding assistance, translation, and decision support. The AMA reports that:
- A strong majority of physicians see clear advantages for patient care.
- Most believe AI can help reduce administrative burden, one of the primary drivers of burnout.
- Optimism about AI’s role in healthcare continues to increase year over year.
The bottom line? Clinicians are not resisting AI. They’re requesting it to eliminate burdens in their existing workflows.
Yet at the American College of Healthcare Executives and other industry forums, the dominant sentiment among digital transformation leaders has been a sort of burnout. In fact, nine out of ten decision-makers report generative AI pilot fatigue. It is no wonder, with the explosion of vendors, the constant stream of “AI-powered” pitches, and the lack of evaluation tools.
The reality is that clinicians may want AI more than their administrators do.
And that gap is where risk, and opportunity, live.
The risk of letting clinicians go it alone
We’ve heard a variety of stories about how clinicians are using unsupported AI in their clinical practice. Most cases boil down to this: Physicians have turned the corner faster than administrators and have experienced the power of AI to reduce administrative work and thereby physician burnout. In fact, the same AMA report listed above found that 73% of Physicians report that they see more and more opportunities for AI to automate administrative tasks that lead to burnout.
But when clinicians choose to use AI that isn’t provided by or sanctioned by their organizations, the risk of data leaving their secure environment increases and governance gaps emerge.
Without sanctioned tools tailor-made for clinical settings:
- Data may leave secure environments.
- Data models may lack validation workflows.
- Outputs may not integrate back into the EHR in structured form.
- Standards become inconsistent.
Where enterprise AI adoption should focus first
In a world saturated with AI vendors, where should health systems get started?
Invisible AI that reduces administrative burden and fits seamlessly into existing workflows.
Not flashy chatbots. Not standalone dashboards. Not experimental pilots disconnected from core systems.
The AMA data makes the priority clear. Physicians are most optimistic about AI when it:
- Eases administrative tasks.
- Reduces documentation time.
- Improves efficiency without adding friction.
That means the first wave of scaled enterprise AI should focus on:
- Patient medical history aggregation.
- Automated chart summarization.
- Structured synthesis of external medical records.
- Ambient documentation integrated into the EHR.
- Inbox triage and message routing.
- Coding and prior authorization support embedded within workflows.
This is what “invisible AI” really means: not that clinicians don’t know it exists, but that it works inside the systems they already use, reducing clicks, cognitive load, and manual work.
When AI forces a new login, a new tab, or a new process, it fails. When AI collapses five steps into one, it sticks
A call to action for digital transformation leaders
Health systems don’t need to chase every AI headline, and they don’t need to move recklessly. But they do need to realize that to meet clinicians where they already are, the old mindset of 2-3 year roadmaps and implementation will leave them behind their competitors. A thoughtful approach might look like this:
1. Start with what clinicians are already trying to solve. Focus early efforts on reducing documentation burden, inbox overload, and manual chart review. These are the areas where physicians already see value, and where the AMA data shows the strongest optimism.
2. Make interoperability a shared priority. When evaluating solutions, emphasize deep EHR integration, structured outputs, and secure data handling. AI should simplify the environment, not fragment it further.
3. Invite clinicians into the evaluation process. Many physicians are already experimenting responsibly. Involving them in vetting and piloting tools not only improves fit, it builds trust and alignment.
4. Provide safe, sanctioned options for high-friction tasks. If clinicians are using AI to summarize lengthy external PDFs or synthesize patient histories, consider how to offer compliant, enterprise-supported tools that address that need directly.
5. Create clear guidance and not barriers. Establish simpler governance frameworks that clarify appropriate use without discouraging innovation. The goal is enablement, not restriction.
This is a new world, where speed and taking on some level of risk are necessary to stay ahead.
Photo credit: Chinnapong, Getty Images
AvaneerDr. Kaushal Kulkarni is board certified ophthalmologist with subspecialty training in neuro-ophthalmology. He is a co-founder and Chief Medical Officer at Predoc: Ax the Fax and Kill The Clipboard and founder and CEO of Eyetamins: consumer eye health which was acquired in 2024.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.
