Good AI is not good enough.
Healthcare leaders have largely moved past the belief that fully autonomous AI can solve their hardest problems. With an estimated 95% of generative-AI initiatives never progressing beyond the pilot stage, they’ve learned that simply “adding AI” does not translate into real-world impact.
Healthcare is no exception. AI is now deeply embedded across the health system, from documentation and scheduling to risk prediction and clinical decision support, forcing the industry to clarify how it’s used. The American Medical Association uses the term “augmented intelligence” to emphasize that AI should help clinicians, not replace them, and stresses that clinical decision-making must still lie with physicians. The American College of Physicians has taken a similar stance, stating that AI should not supplant physician judgment.
Physicians agree. Surveys show they want AI to lighten administrative load and strengthen care quality, not act autonomously. The goal is not to build systems that operate on their own. The emerging model is hybrid intelligence as a force multiplier that combines advanced technology with human insight, oversight, and accountability.
Why Leaders Fear Autonomous AI
In a national survey, health system leaders were asked about their perceptions of AI model types, adoption barriers, and the role of human oversight in clinical workflows. Respondents consistently indicated that autonomous, “black box” AI is risky and insufficient for high-stakes clinical environments, while AI paired with expert human validation is viewed as safer and more accurate.
- 62.5% identified “misinterpretation of data” as the top risk when AI operates without human oversight.
- Only 12.5% said autonomous AI has delivered meaningful value to their work to date.
By contrast:
- 75% rely on expert human validation to ensure AI output is clinically relevant.
- 75% rated clinician involvement in AI design and deployment as “critically important.”
Healthcare leaders aren’t looking for a magic switch to automate healthcare. Instead, they want a force multiplier. They value solutions that respect clinical expertise, demand human validation, and integrate seamlessly into existing workflows. The winning strategy in this sector is not to build a better machine, but to build a better team, one where AI provides the speed and scale, and humans provide the judgment and care.
Hybrid Intelligence Is Winning
A recent Lancet Digital Health study shows why this model works. Researchers tested five leading generative AI models against physicians solving complex diagnostic cases from Massachusetts General Hospital. Although the highest-performing model outscored individual resident physicians, the most significant gains appeared when humans and AI worked together.
When physicians reviewed the model’s ranked differential diagnoses, their accuracy improved dramatically, sometimes nearly doubling. Their diagnostic lists became more complete without losing the reasoning that guides patient care. The model surfaced possibilities clinicians might not have considered, while clinicians provided the context the model lacked.
The collaboration worked in the other direction, too. When the physicians’ differentials were fed back into the models, the models themselves became more accurate.
In other words, AI sharpened human thinking, and human thinking sharpened AI. It’s the force-multiplier effect in action.
Where Hybrid Intelligence Delivers the Greatest Impact
The greatest hybrid intelligence opportunities lie in workflows that require both accuracy and throughput. Clinical documentation, diagnostic support, care coordination, and quality measurement all meet that description. But nowhere is the need more visible than in clinical data abstraction.
Clinical data abstraction in hospitals is the process of clinicians manually reviewing a patient’s electronic medical record to answer very specific questions for clinical registries. These registries are national, standardized databases that track patients with similar conditions or procedures, and they are essential for quality measurement, process improvement, and regulatory reporting. Specifically, they help hospitals track outcomes, assess treatments, refine care pathways, and demonstrate adherence to established standards.
Unfortunately, current methods of manual data abstraction are time-consuming, labor-intensive, costly, and prone to human error. Health systems in the U.S. spend between $10B and $15B annually to manually abstract data.
AI can surface the most relevant elements of a record, identify missing or conflicting information, and present a structured starting point for abstraction. Clinicians then validate, correct, and refine the AI-generated outputs. They apply their expertise to ensure the final data is accurate and consistent with clinical judgment.
That method mirrors the dynamic shown in the Lancet Digital Health study. AI expands the field of possibilities and accelerates the process, and clinicians provide the context and oversight.
Healthcare workflows that demand precision at scale, like clinical registries, documentation review, and diagnostic reasoning, can’t rely on automation alone. The solutions that last are the ones that are built around partnership with clinicians, within workflows, and inside the operational realities of healthcare.
The goal isn’t to automate the clinician out of the process. It’s to design force-multiplier workflows where clinicians can do their best work, supported by technology that makes that work faster, safer, and more accurate.
About Brent Dover
Brent Dover is the CEO at Carta Healthcare, the leader in enterprise clinical data management. With a deep commitment to improving healthcare outcomes, Brent leads Carta Healthcare’s mission to unlock the power of clinical data for hospitals, health systems, and life sciences organizations. His strategic vision drives the company’s focus on leveraging advanced artificial intelligence and expert clinical professionals to streamline data abstraction, enhance operational efficiency, and generate actionable insights that ultimately elevate patient care. Under his leadership, Carta Healthcare has experienced significant growth and recognition for its innovative approach to solving complex healthcare data challenges.
