Augmented intelligence in medicine: what the AMA is doing and why it matters
Augmented intelligence has become a buzzword in modern medicine, and for good reason.
While the debate over the role of artificial intelligence in healthcare heats up, the American Medical Association — the largest medical organization in the United States — has already made its position clear: technology should expand human capability, not take the wheel.
And honestly, that stance matters way more than it might seem at first glance.
The speed at which AI tools are making their way into clinics, hospitals, and medical offices around the world is staggering. In just three years, the number of physicians using AI on a daily basis has practically doubled, and today more than 80% of professionals report some form of these technologies in their workflow.
That rapid growth has brought along a set of questions no one can afford to ignore anymore:
- Who is responsible when AI gets it wrong?
- How do you protect patient data?
- Can the doctor-patient relationship survive this transformation?
This is exactly where the conversation around healthcare ethics comes in full force — and where the AMA’s work starts making even more sense. Let’s dig into what’s happening in this space, what the data shows, and where this path seems to be heading. 🩺
What the AMA means by Augmented Intelligence
When the AMA talks about augmented intelligence, they’re not talking about robots replacing doctors or algorithms making clinical decisions on their own. The concept is quite different from that — and a lot more interesting. The core idea is that artificial intelligence works as an extra layer of support, processing massive volumes of data, identifying patterns the human eye might miss, and delivering that information in an organized way so the healthcare professional can make the final call with much more context and confidence.
The AMA’s own House of Delegates formally adopted the term augmented intelligence specifically to reinforce this conceptual difference. This isn’t just a matter of semantics: by choosing that label, the organization signals to the market, regulators, and the professionals themselves that the design of these tools needs to start from the premise that the human stays at the center of the decision. Think of it like having an incredibly fast, attentive, tireless assistant — but one that still needs a human to validate every step.
This model of human-machine partnership makes especially good sense in medicine, where the consequences of a mistake can be severe and irreversible. An AI system can analyze thousands of imaging studies in minutes, highlight anomalies, compare them against global databases, and present diagnostic hypotheses — but the person who looks the patient in the eye, talks with them, understands the context of their life, and signs off on the treatment decision is still the physician. That balance is, according to the AMA itself, not just desirable but essential for technology to be truly useful and safe in the clinical setting.
The organization also argues that AI development aimed at healthcare needs to be built with this philosophy from the ground up — meaning it’s not something you tack on later as an afterthought. When systems are designed to amplify the professional rather than replace them, the chances of critical errors go down, physician adoption goes up, and most importantly, patient trust in the process is preserved.
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.
AMA policies for AI development and use in healthcare
The AMA hasn’t stayed in the conceptual lane alone. The organization has developed a robust set of policies addressing the development, deployment, and use of AI in healthcare settings. These guidelines cover topics ranging from clinical oversight of tools to transparency in the use of these technologies, along with generative AI governance, physician liability, data privacy, cybersecurity, and even the use of AI by health insurers in automated decision-making systems.
The central points of these policies include:
- AI oversight in healthcare — ensuring there is a governance structure to monitor the performance of tools in use
- Transparency — defining when and what should be communicated to both physicians and patients about AI use
- Generative AI governance — creating specific guidelines for models like natural language ones that are rapidly entering the clinical routine
- Physician liability — clarifying the boundaries of a professional’s responsibility when using AI-enabled technologies
- Privacy and cybersecurity — protecting sensitive patient data against breaches and misuse
- Use by health insurers — regulating how insurance companies can employ AI and automated systems in approving or denying procedures
That last point is particularly relevant and sparks plenty of discussion in the United States. When a health insurer uses an algorithm to automatically decide whether a treatment will be covered or not, the patient can be harmed without even knowing the decision was made by a machine. The AMA has taken a firm stance in favor of transparency in these cases and advocates that physicians need an active voice in building those rules.
What physicians think about AI — the data from 2023 to 2026
In 2023, the AMA conducted a comprehensive study with more than a thousand physicians to understand how they view the use of AI in healthcare. The survey assessed current use, motivations for future adoption, key concerns, areas of greatest opportunity, and implementation requirements. Given how quickly the landscape evolves, the study was repeated at the end of 2024 and again in 2026.
The results show a clear trend of growth in both adoption and confidence:
- More than 80% of physicians report using AI in their professional work in 2026 — double what was recorded in 2023
- More than three-quarters of physicians say AI improves their ability to care for patients, up from 65% in 2023
- About 40% of physicians say they feel both excitement and concern about AI’s role in healthcare at the same time
The top concerns continue to be protecting patient privacy and preserving the integrity of the doctor-patient relationship. These two points consistently appear at the top of the worry list, regardless of specialty or the professional’s age group.
In 2026, the study was expanded to include two new dimensions: physicians’ perspectives on patients using AI themselves and professional training needs, including concerns about the potential loss of clinical skills as AI adoption grows. That last point is fascinating — and a little unsettling. If physicians become too dependent on automated tools for tasks like differential diagnosis or exam interpretation, there is a real risk that certain clinical competencies could weaken over time. It’s a question that deserves ongoing attention.
Healthcare Ethics at the Center of the Debate
With the rapid expansion of artificial intelligence in medicine, ethical questions have moved from academic discussion to practical urgency. One of the most sensitive points is accountability for errors. When an algorithm contributes to a misdiagnosis, who answers for it? The physician who used the tool? The company that developed the system? The hospital that implemented it? This chain of responsibility still doesn’t have a clear answer in most countries, and as long as that gap exists, healthcare ethics remains unstable ground for everyone involved — professionals, patients, and developers.
Another point that raises red flags is patient data privacy. AI systems in medicine are trained and fed with extremely sensitive information — medical histories, test results, genetic predispositions, behaviors, and even financial data when it comes to health insurance. The misuse of that information, whether through leaks, unauthorized sales, or improper access, represents a real threat to people’s dignity and safety. Regulations like LGPD in Brazil and HIPAA in the United States try to create barriers, but the pace of technological advancement frequently outpaces the speed of legislation.
There’s also a dimension that tends to fly under the radar but is equally important: algorithmic bias. Artificial intelligence models learn from historical data, and if that data reflects inequalities in the healthcare system — like the underrepresentation of certain population groups in clinical studies or disparities in access to quality diagnostics — the algorithm will reproduce and even amplify those inequalities. Healthcare ethics in the context of AI therefore demands an active commitment to equity from the very design phase of these systems.
The AMA’s AI Specialty Collaborative
To ensure that physicians play a central role in how AI is developed and integrated into healthcare, the AMA created the AI Specialty Collaborative. This initiative brings together 21 medical specialty societies with a shared goal: making sure the voice of the professionals who are on the front lines of care is heard in the decisions that will shape the future of these technologies.
The logic behind this collaborative is simple and powerful. Every medical specialty has its own particularities — what works in radiology may not make sense in cardiology, and the risks of an AI tool in dermatology are different from the risks in oncology. By bringing these different perspectives together, the AMA can build guidelines that are more comprehensive, realistic, and applicable, instead of producing generic guidance that doesn’t end up serving anyone well.
AI in medical education
Artificial intelligence is also gaining ground across every stage of medical training, both as a tool for educators and students and as a subject of study in its own right. The AMA sees AI as a key component of what’s called precision education — a model that personalizes the learning process according to each student’s needs and pace — as well as precision health, which tailors clinical care to the individual patient’s profile.
This dual dimension matters because it prepares future physicians not just to use the tools that already exist, but to understand their principles, their limitations, and their potential for evolution. A professional who understands how a machine learning model works is much better equipped to critically evaluate the results the tool delivers — and that’s exactly the kind of competency the augmented intelligence concept demands.
How AI Development is Shaping Medicine
AI development focused on medicine has advanced in areas that until recently seemed far removed from clinical reality. Image analysis tools can now detect tumors in X-rays and CT scans with a level of accuracy that, in some studies, matches or surpasses that of experienced specialists. Natural language processing systems are being used to analyze medical records, identify patterns across large populations, and even suggest adjustments to treatment protocols based on evidence updated in real time.
In the field of genomics, artificial intelligence is transforming precision medicine. Algorithms can cross-reference genetic profiles with global databases and identify which treatments are most likely to work for a given patient, reducing the trial-and-error process that is still very common in oncology, for example. This not only improves clinical outcomes but also cuts down on patient suffering and treatment costs.
But AI development for healthcare doesn’t happen in a vacuum — and that’s where the critical role of organizations like the AMA comes in. Without clear guidelines, without clinical validation standards, and without a rigorous approval process, any company can launch a tool on the market promising to revolutionize medicine but without the consistency needed to be used safely in real-world settings. Regulating the development of these technologies is, therefore, just as important as the innovation itself. 🔬
Recent updates and key milestones
The pace of developments in healthcare AI has been intense, and the AMA has kept up — and in many cases led — these movements. Among the most notable updates:
- In October 2025, the AMA launched the Center for Digital Health and AI, a dedicated structure designed to put physicians at the center of the process of shaping, guiding, and implementing AI tools and other technologies that are transforming medicine
- In May 2026, the AMA published a patient-facing infographic with guidance on how to navigate AI in healthcare safely and effectively
- The AMA publicly weighed in on the U.S. federal government’s AI action plan, launched in 2025, expressing willingness to collaborate with the administration on key areas of regulation and implementation
- A paper published in the Journal of Medical Systems, titled Trustworthy Augmented Intelligence in Health Care, consolidated the findings of the AMA’s literature review on the challenges AI brings to the sector and the practical guidance available
- A report from the organization provides an overview of state legislative activity and discusses three priority areas: use of AI by health insurers, transparency, and physician liability
These initiatives show that the AMA isn’t just reacting to events but actively positioning itself to influence the rules of the game while they’re still being written.
CPT codes and the classification of AI in medical practice
A highly technical but enormously impactful aspect is the integration of AI into the CPT (Current Procedural Terminology) code system. This system is the backbone of healthcare communication in the United States — it’s through CPT codes that medical procedures and services are recorded, processed, and reimbursed.
The AMA provides resources and guidance on how the updated CPT code set classifies different AI applications, and it also has the Digital Medicine Payment Advisory Group, known as DMPAG. This group identifies barriers to digital medicine adoption and proposes comprehensive solutions related to coding, payment, coverage, and other topics. It might sound like a bureaucratic detail, but without this kind of infrastructure, AI tools simply can’t be sustainably integrated into the actual workflow of physicians and healthcare institutions.
The Doctor-Patient Relationship in a World with AI
One of the most deeply human concerns amid all this technological transformation is what happens to the relationship between doctor and patient when artificial intelligence enters the room. There’s a legitimate fear — especially among older patients or those less familiar with technology — that the visit will become a cold interaction mediated by screens and algorithms, losing that element of listening, empathy, and presence that is a fundamental part of care.
On the other hand, when augmented intelligence is used thoughtfully, it can actually free physicians to dedicate more time to the human side of the visit. If an AI system has already processed the patient’s history, flagged points of concern, and organized the most relevant information before the doctor even walks in, the professional arrives at the appointment with much more context — and can use the available time to have a real conversation with the patient, understand their questions, their fears, and their expectations, instead of spending precious minutes reviewing documents or entering data.
What will determine which of these scenarios plays out isn’t the technology itself, but the way it’s introduced into healthcare practices — and that’s where professional training, institutional policies, and industry guidelines make all the difference. Physicians who understand what AI does, what its limits are, and how to use it in the patient’s favor are in a much better position to get the most out of this partnership. Likewise, patients who are informed about how their data is used and how technology supports — rather than replaces — the professional’s judgment tend to trust the process more and engage better in their own care. Transparency, in this context, isn’t just an ethical matter: it’s a therapeutic tool. 💙
Educational resources and continuing education
The AMA also invests heavily in continuing education for physicians who need to navigate AI’s increasingly central role in clinical practice. The AMA STEPS Forward program offers a collection of digital health solutions that help professionals integrate AI into workflows, reduce administrative burden, and improve patient care — always with attention to issues like ethics, bias, and professional well-being.
These resources include case studies, implementation strategies, and expert perspectives, along with webinars on AI governance, documentation tools, and governance frameworks. There are also podcasts featuring hands-on experiences from major health systems like The Permanente Medical Group and Johns Hopkins, which have already implemented robust AI programs.
On AMA Ed Hub and the JAMA network, physicians can access content on the components of AI in healthcare, explore challenges and opportunities, and earn continuing medical education credits. This learning ecosystem is critical because technology evolves so fast that relying solely on what was taught in medical school is no longer enough — staying current needs to be ongoing and accessible.
Where this path is heading
The current landscape makes it clear that artificial intelligence in medicine isn’t a passing trend — it’s a structural transformation happening right now that will only deepen in the years ahead. Global investments in healthtech with an AI focus keep growing, new models are launched regularly, and clinical applications approved by regulatory bodies like the FDA in the United States already number in the hundreds. The pace isn’t going to slow down, and that means the decisions being made today — about how to regulate, how to train professionals, how to protect data, and how to ensure equity — will define how this chapter of medicine is remembered.
The concept of augmented intelligence championed by the AMA represents, in this context, an important compass. It offers a north star that is both ethical and practical at the same time: technology as an ally, the human as the protagonist, and responsibility shared among professionals, developers, and institutions. This model isn’t perfect — none is — but it starts from a premise that puts patient well-being at the center, and that alone is a huge differentiator in a field where tech hype sometimes moves faster than the ability to assess consequences.
The future of medicine with artificial intelligence will be built in the details: in the public policies that define how data can be used, in the medical school curricula that teach future doctors to work with these tools, in the AI development decisions that prioritize rigorous clinical validation before launch, and in the honest conversations about healthcare ethics that need to involve not just experts but also the patients who will be affected by these technologies. The combination of technological advancement and human responsibility is what will separate a genuine revolution from a large-scale failed experiment. 🚀
