Augmented intelligence in medicine: how the AMA is shaping the future of AI in healthcare
Augmented intelligence is the term the American Medical Association (AMA) prefers to use instead of artificial intelligence, and that choice is no accident. The decision reflects a very clear vision about the role of technology in modern medicine: it didn’t come to take the doctor’s seat, but to make that professional even more capable of caring for those who need it.
The AMA House of Delegates formally adopted the concept of augmented intelligence as a way to emphasize that the design of these technologies should enhance human intelligence, not replace it. This conceptual distinction may seem subtle, but it has profound implications for how physicians, patients, and developers approach the use of AI in clinical settings.
And the numbers show this transformation is already happening at a rapid pace. In 2023, about 40% of physicians reported using some AI-powered tool in their daily practice. Today, in 2026, that number has surpassed 80% 🚀 That’s double in less than three years. Trust in these tools is also growing: more than three-quarters of physicians say AI improves their ability to care for patients, a significant jump from the 65% recorded in 2023.
But alongside the enthusiasm comes a healthy dose of caution — after all, when we’re talking about healthcare, mistakes carry a different weight. About 40% of physicians say they feel both excitement and concern about AI’s role in healthcare. Issues like patient privacy, ethics in data use, and preserving the integrity of the doctor-patient relationship are at the center of the debate the AMA has been leading with considerable seriousness.
In this article, we’ll explore how all of this is unfolding, what physicians themselves think about it, what policies are being created, and which concrete initiatives are shaping the future of AI in medicine. 🩺
Why augmented intelligence and not artificial intelligence?
The difference between the two terms might seem small, but it carries a very specific worldview. When the AMA adopts the concept of augmented intelligence, it’s signaling that technology’s role in healthcare is to enhance human judgment, not replace it. This matters a lot in clinical practice because the physician remains ultimately responsible for every decision made about a patient, regardless of what any artificial intelligence system may have suggested.
Technology enters as a support layer, capable of processing enormous volumes of data, identifying patterns, and presenting options — but always leaving clinical reasoning and human empathy in the driver’s seat. Computer vision algorithms can help radiologists detect abnormalities in imaging exams, large language models can assist with clinical documentation, and decision support systems can suggest evidence-based treatments. However, in all these scenarios, the physician remains the final decision-maker.
This stance also has an important effect on how healthcare professionals relate to these tools. When the narrative is about replacement, physicians tend to see AI as a threat or as something that diminishes their role. When the narrative is about enhancement, the same technology becomes an ally that frees up time for what truly matters: looking patients in the eye, listening to their stories, understanding the life context that data will never fully capture. This shift in perspective is subtle, but it has a direct impact on the adoption and responsible use of these solutions in clinics and hospitals.
Beyond that, the technical advancement of artificial intelligence systems applied to medicine itself demands this conceptual clarity. Each tool has very specific capabilities and limitations. Understanding that they augment human capacity rather than acting autonomously helps physicians, administrators, and patients maintain more realistic expectations and use these solutions in a smarter, safer way.
The AMA’s policies for AI development, deployment, and use
The AMA isn’t just watching AI’s growth in medicine from the sidelines. The organization has been actively committed to ensuring that artificial intelligence reaches its full potential to advance clinical care and improve the well-being of healthcare professionals. With the growing number of AI-enabled tools in healthcare, it’s essential that they be designed, developed, and deployed in an ethical, equitable, and responsible manner.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
A central point of these policies is transparency. The use of AI in healthcare needs to be transparent for both physicians and patients. Beyond medical devices, AI is increasingly being used in hospital administration and in reducing the bureaucratic burden on physicians. Recognizing that policies and guidance are needed for both device and non-device applications, the AMA has developed new guidelines addressing fundamental aspects:
- AI oversight in healthcare — who monitors and who is accountable for outcomes
- Transparency — when and what to disclose to patients and professionals
- Generative AI policies — specific governance for language models and generative tools
- Physician liability — how to define physician responsibility when using AI-enabled technologies
- Data privacy and cybersecurity — protection of sensitive patient information
- AI use by health insurers — regulation of automated decision-making systems
These guidelines represent a concrete effort to create a framework that allows innovation without compromising safety and patient rights. The AMA Board of Trustees is responsible for implementing these policies, and given the number of stakeholders and policymakers involved in the evolution of AI in healthcare, the organization recognizes the importance of not only adopting a baseline set of guidelines but continuing to refine them as the technology develops.
What do physicians really think about all this?
The raw adoption numbers are impressive, but they don’t tell the full story. The AMA conducted a comprehensive study in 2023 with more than 1,000 physicians about their perceptions regarding AI use in healthcare, including current use, future motivations, key concerns, areas of greatest opportunity, and requirements for adoption. Given the rapid pace of evolution in the AI landscape, the study was repeated in late 2024 and again in 2026.
The objectives of this research are clear:
- Capture sentiment among practicing physicians regarding the increasing use of AI in healthcare
- Evaluate AI use cases based on familiarity, relevance, and usefulness
- Identify key resources and areas of need so physicians can consider implementing AI tools in their practice
In 2026, the study was expanded to examine two additional areas: physicians’ perspectives on AI use by patients and training needs for professionals, including concerns about the potential loss of clinical skills as AI adoption grows. That last point is especially interesting because it touches on something many professionals feel but rarely discuss openly — the fear that technology dependence could weaken competencies that took years to develop.
The leap from roughly 40% to over 80% adoption shows that artificial intelligence has gone from a curiosity to a part of daily medical practice. But a good number of professionals still carry legitimate questions about how these tools work behind the scenes. Many report using AI-based applications without fully understanding how results are generated, which creates a real tension between technological convenience and clinical responsibility.
This scenario raises a very important question about trust. For a physician to trust a suggestion generated by an AI system, they need access to clear information about how that system was trained, what data was used, which populations are represented in that data, and where the margins of error are greatest. Without that clarity, there’s a risk that the professional will start delegating decisions to the tool without the necessary critical thinking — or, at the other extreme, will dismiss useful suggestions due to a lack of trust.
On the other hand, physicians who report positive experiences with augmented intelligence tend to highlight the same benefits: more time with patients, faster diagnoses in complex cases, and a sense of confidence from having a second analytical perspective on data that would be difficult to process manually. These accounts show that when the technology is well-implemented, well-explained, and well-integrated into the clinical workflow, it truly fulfills the role the AMA advocates for. 💡
The AMA’s AI Specialty Collaborative
One of the most relevant initiatives in this ecosystem is the AMA’s AI Specialty Collaborative, which brings together 21 medical specialty societies. The goal is to ensure that physicians play a central role in how AI is developed and integrated into healthcare practice. This multidisciplinary collaboration is essential because each medical specialty has different needs, workflows, and risks. What works for dermatology may not apply in the same way to cardiology or oncology.
By bringing these diverse voices to the table, the AMA can build guidelines that are both comprehensive and specific enough to be applicable in the real-world practice of each field. This avoids the trap of creating regulations that are too generic to serve anyone, or so specific that they become impractical outside a single context.
AI in medical education
Artificial intelligence is playing an increasingly important role across every stage of the medical education continuum, both as a tool for educators and students and as a subject of study in its own right. AI has the potential to transform the educational experience as part of what’s called precision education, and to transform patient care as part of precision health.
This means that future physicians are already being trained in an environment where AI is part of the reality. They learn to use these tools critically, understand their limitations, and develop a more natural relationship with the technology from the very beginning of their training. This generational shift is likely to further accelerate the responsible integration of AI into medical practice in the coming years.
Ethics and transparency as non-negotiable pillars
When it comes to healthcare, ethics isn’t a regulatory detail — it’s the foundation of everything. The discussion about using artificial intelligence in medicine had to mature quickly, especially because the data involved is among the most sensitive there is: clinical history, genetics, behavior, chronic conditions. Information that, in the wrong hands or poorly managed, can cause serious harm to people’s lives.
The AMA has been quite vocal in advocating that any application of AI in medicine must respect clear principles of privacy, informed consent, and non-discrimination — and that these requirements need to be part of the design of these solutions from the start, not an adjustment made after problems surface.
Transparency enters this context as both a technical and a human component. From a technical standpoint, it means that augmented intelligence systems need to be auditable — it should be possible to understand how a given recommendation was generated, which variables weighed most heavily in that result, and where the model may be operating with less precision. From a human standpoint, it means physicians and patients need to know when they’re interacting with an AI-based solution, what it can and cannot do, and how to challenge or override its suggestions when clinical judgment points in a different direction.
The AMA also published an article in the Journal of Medical Systems titled Trustworthy Augmented Intelligence in Health Care, presenting findings based on a literature review of the challenges AI in healthcare presents, along with reflections on existing guidance. This kind of applied academic work is essential for translating theoretical debates into practical guides that physicians and administrators can actually use in their day-to-day work.
Recent updates and milestones
The AMA has been working with technology and healthcare leaders to bring insights about the potential applications of AI and ensure that physicians have an active voice in this process. Some of the most recent developments are worth highlighting:
- In May 2026, the AMA released an infographic to help patients navigate the use of AI chatbots in healthcare safely and effectively
- In October 2025, the Center for Digital Health and AI was launched — a center dedicated to placing physicians at the heart of the development, guidance, and implementation of AI tools and other technologies transforming medicine
- The AMA welcomed the U.S. federal government’s AI action plan in 2025, expressing willingness to work with the administration on key areas of AI regulation, policy, and implementation
- An AMA issue brief provides an overview of recent state legislative activity and discusses three key areas of AI policy: use by health insurers, transparency, and physician liability
These milestones show that the discussion about AI in medicine isn’t stuck in abstract debates. Concrete actions are being taken on multiple fronts — educational, regulatory, technological, and political — that together are building the framework needed for responsible and beneficial adoption.
CPT and the coding of AI in medical practice
One aspect that often goes unnoticed in discussions about AI in healthcare is the issue of coding and reimbursement. The CPT (Current Procedural Terminology) code set is the standard that drives communication across the healthcare system, enabling the processing and advanced analysis of medical procedures and services.
The AMA offers resources to guide the classification of various AI applications within the CPT system, as well as specialized consulting through the Digital Medicine Payment Advisory Group (DMPAG). This group identifies barriers to digital medicine adoption and proposes comprehensive solutions covering coding, payment, coverage, and more. Without this classification and reimbursement infrastructure, many AI tools simply couldn’t be sustainably integrated into medical practice because there would be no way to adequately compensate the professionals who use them.
Educational resources and continuing education
The AMA also recognizes that physicians need support to navigate the rapidly evolving role of AI in clinical practice. The AMA STEPS Forward program offers a collection of digital health solutions that help provide insights into integrating AI into workflows, reducing administrative burden, and improving patient care — all while addressing issues like ethics, algorithmic bias, and physician well-being.
These resources include case studies, implementation strategies, and expert perspectives, equipping physicians with the knowledge and tools to adopt AI responsibly and effectively. They are open-access materials eligible for continuing medical education credits, which encourages professionals to stay up to date without needing to invest time and money in complex supplementary training programs.
The AMA Ed Hub also offers specific courses on AI in healthcare, and the JAMA Network provides curated article collections on machine learning. This educational ecosystem matters because the speed of AI innovation demands that healthcare professionals stay in constant learning mode, and having reliable, accessible sources makes all the difference in that process. 📚
What to expect in the near future
With more than 80% of physicians already using some form of artificial intelligence in their daily practice, the debate is now shifting from whether to adopt to how to adopt better. The next steps involve an increasingly deep integration between clinical systems, electronic health records, and AI-based decision support tools — which will require more robust technical standards for interoperability and, at the same time, more detailed ethical protocols to ensure this integration doesn’t open gaps for the misuse of patient data.
Augmented intelligence is also expected to become increasingly personalized. Models that currently operate with large population-level data sets are likely to evolve into systems capable of more precisely accounting for each patient’s unique characteristics: their individual history, their preferences, their specific conditions. This has enormous potential to improve the quality of care, but it also raises new questions about ethics and transparency, especially when it comes to deciding which data feeds these systems and who has access to the conclusions they generate.
The AMA’s 2026 survey has already begun exploring these territories, including physicians’ perspectives on AI use by patients themselves. This is a fascinating angle because it highlights a new challenge: how does a physician handle a patient who arrives at the office with diagnoses or suggestions generated by AI tools they used on their own? This dynamic is going to redefine the professional-patient relationship in ways that are still being understood.
What seems clear, looking at the trajectory the AMA and other organizations are charting, is that the future of AI in healthcare won’t be built by engineers and data scientists alone. It will depend on engaged physicians, informed patients, attentive regulators, and an institutional culture that places ethics at the center of every technological decision. Technology has the potential to transform medicine for the better, and that potential is fully realized when it’s developed and used with responsibility, clarity, and genuine respect for human life. 💙
