Redefining surgical expertise in the age of artificial intelligence

Key points

Variability has long been an accepted feature of surgical practice, driven by differences in surgeons’ training, case exposure, and individual technical aptitude. Traditionally, objective methods to assess surgical skill have been limited to indirect measures such as reported outcomes and case volume (1). Artificial intelligence (AI) has the potential to create a new dimension wherein technique could be directly linked with the gamut of objectively assessed outcomes. Through continuous data capture and analysis, AI systems may enable granular, objective assessment of technical performance, decision-making, and team communication, along with the opportunity to link these variables to real-world patient outcomes. In doing so, they offer a path toward reducing variability and standardizing surgical care at an unprecedented level. In the near term, this may translate into greater consistency in how procedures are performed among surgeons and across institutions. As systems trained on datasets and optimized using collaborative learning frameworks establish consistent protocols, autonomous robotic surgery is poised to become a reality.

By minimizing variability towards a mean, an AI-driven path helps deliver consistent patient care, a direction the field should broadly support. However, this approach may inadvertently effectuate risk aversion, suppressing creativity and preventing or discouraging surgeons from deviating from established norms. Such an environment could limit new approaches and constrain surgical progress. Innovation in surgery has historically been driven by human creativity and experiential insight, not algorithmic optimization. Although AI is effective at enhancing and expanding established methodologies, it relies on pre-existing datasets and lacks the capability to independently develop new paradigms that serve as the foundation for such datasets. The concern, therefore, is not that AI will fail to improve consistency, but that its success may, as a downstream consequence, come at the expense of innovation.

In their lead article, Granados et al. envision a future where autonomous robotic systems are seamlessly integrated into the operating room, handling setup, real-time monitoring, workflow feedback, scheduling optimization, technical assistance, and eventually performing procedures under human supervision (2). Early demonstrations of autonomous surgical capability have already been described in controlled settings (3), highlighting the feasibility of this trajectory. This transformation will not occur abruptly. Integration will evolve with early adoption focused on lower-risk, high-efficiency domains that improve workflow (e.g., patient positioning, instrument reconciliation, and operating room logistics). As these systems demonstrate reliability and generate trust, AI will be incorporated into decision support, selective technical assistance, and finally supervised operative autonomy.

However, the “art” of surgery extends far beyond technical execution. While operative skill remains essential, the true artistry of surgery lies in judgment: deciding when to operate, whom to operate on, and how to execute the appropriate procedure. These decisions are nuanced, patient-specific, and shaped by experience and the clinical context. As autonomous systems become integrated into surgical care, there is a growing possibility that AI will not only guide how a procedure should be performed, but also begin to inform the more important questions of why it should be performed and on whom. While such capabilities may enhance decision-making through data-driven insights, they also challenge one of the most fundamental roles of the surgeon: serving as the primary liaison of clinical judgment. AI will not replace clinical judgment outright but will gradually redefine it, shifting surgeons from decision-makers to validators of decisions.

In this vision of the future operating room, the surgeon transitions from technical expert to supervisor of increasingly autonomous systems. In this model, the surgeon’s primary responsibility becomes one of oversight: determining whether the actions and outputs of these systems are appropriate at each step of care. However, this raises an underexplored challenge. The ability to judge whether an action is “right” or “wrong” in surgery is not innate—it is developed through years, often decades, of direct operative experience, independent decision-making, and the management of complications. If AI systems progressively assume both the technical and cognitive burdens of surgery, opportunities for trainees and early-career surgeons to develop this expertise may diminish. Prior work in human factors has shown that increasing reliance on automation can degrade situational awareness and erode human performance over time (4).

This creates a fundamental paradox: if mastery is required to supervise AI systems, but the pathway to mastery is diminished by those same systems, the long-term sustainability of the surgical profession comes into question. While AI promises to improve consistency, efficiency, and outcomes, its integration must be carefully balanced against the need to preserve the experiential foundation upon which surgical expertise is built. At the same time, it is worth acknowledging a more radical possibility: this paradox itself may ultimately become irrelevant if AI-driven surgical systems evolve to such a level of reliability that human oversight is no longer necessary. In such a future, where outcomes are consistently optimized beyond human capability, the need for traditional surgical mastery may diminish entirely. Whether such a transformation is one that the profession and society are prepared to embrace remains uncertain.

References

• 1

  BirkmeyerJDFinksJFO’ReillyAOerlineMCarlinAMNunnARet al. Surgical skill and complication rates after bariatric surgery. N Engl J Med (2013) 369:1434–42. doi: 10.1056/NEJMsa1300625

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 2

  GranadosAKhannaRFischerNRaisonNCiabattiniMRobertshawHet al. Evolving surgical teams in the age of artificial intelligence and robotics. Front Sci (2026) 4:1783803. doi: 10.3389/fsci.2026.1783803

  • CrossRef
  • Google Scholar

• 3

  ShademanADeckerRSOpfermannJDLeonardSKriegerAKimPC. Supervised autonomous robotic soft tissue surgery. Sci Transl Med (2016) 8:337ra64. doi: 10.1126/scitranslmed.aad9398

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4

  EndsleyMR. From here to autonomy: lessons learned from human–automation research. Hum Factors (2017) 59(1):5–27. doi: 10.1177/0018720816681350

  • Pubmed Abstract
  • CrossRef
  • Google Scholar