SENTPO CAREER GUIDE | AI AND HEALTHCARE

Will AI Replace Medical Doctors by 2036? Future Skills, Career Changes and Courses to Consider

Artificial intelligence is changing how healthcare professionals analyse medical information, document consultations, interpret medical images and support clinical decisions. Discover how medicine could evolve over the next 10 years, which skills doctors should develop, and which AI courses medical students and practising physicians can consider.

Medicine has always evolved alongside scientific and technological progress. Advanced imaging, electronic health records and robotic-assisted procedures have already changed how doctors investigate diseases and deliver care. Artificial intelligence is another important development that medical professionals need to understand.

For students considering medicine, this raises an important question: will becoming a doctor still be a good career choice in 2036?

For practising doctors, another question matters just as much: what should they learn now to remain effective and adapt to an increasingly AI-enabled healthcare system?

This guide distinguishes current developments from possible future changes. No reliable evidence establishes exactly how many doctors AI will replace by 2036. The impact will depend on clinical evidence, regulation, healthcare infrastructure, costs and patient acceptance.

1. How Is AI Already Changing Medicine?

AI in healthcare is not merely a future concept. It is already being used in selected clinical, research and administrative applications.

The US Food and Drug Administration maintains a public list of AI-enabled medical devices authorised for marketing in the United States. These include technologies for particular applications in radiology, cardiovascular care and other clinical areas. Authorisation applies to specific products and intended uses; it does not mean that every AI tool is accurate or suitable for every patient.

Explore the FDA’s AI-enabled medical device information →

The American Medical Association’s 2026 physician survey reported that 81% of surveyed US physicians used AI professionally. Reported applications included summarising medical research, preparing clinical documentation, creating discharge instructions and summarising patient charts. This finding describes surveyed US physicians, not a global adoption rate or a forecast of job losses.

Read the AMA physician survey →

Medical imaging and diagnostic support

AI-enabled systems can assist with analysing particular medical images, identifying patterns and flagging findings for review. Depending on the technology, they may help clinicians prioritise cases or assess specific abnormalities.

Clinical documentation

AI tools can help draft clinical notes, organise information and summarise consultations. Doctors must review the output for omissions, incorrect statements and clinically important details.

Medical research and evidence review

AI can help doctors navigate medical literature and summarise research. However, summaries and references need verification because AI systems can misrepresent findings, overlook study limitations or produce inaccurate information.

Clinical decision support

Some systems analyse patient information and provide alerts, risk estimates or recommendations. Their usefulness depends on data quality, clinical context and the system’s validated intended use.

Patient communication and administration

AI may help prepare draft discharge instructions, patient education materials and routine communications. Doctors must ensure that information is accurate, suitable for the individual patient and understandable.

Healthcare research and drug development

AI is being explored in biomedical research, including the analysis of biological data and the development of potential treatments. A promising AI result does not replace scientific validation, clinical trials or applicable regulatory processes.

2. Six Ways Medicine Could Change Between 2026 and 2036

The following are plausible developments, not guaranteed predictions. Adoption will vary by speciality, country, institution and healthcare system.

Change 1: AI-assisted diagnosis may become more common

AI systems may increasingly help doctors identify patterns, organise differential diagnoses and highlight information that requires further investigation.

For example, a physician evaluating a patient with several symptoms may use an appropriate clinical decision-support tool to organise possible explanations and locate relevant guidance. The doctor must still consider medical history, examination findings, risk factors and test results.

What to learn: Clinical reasoning, AI literacy, diagnostic accuracy assessment, critical evaluation of recommendations and management of uncertainty.

Change 2: Medical documentation may become increasingly automated

AI tools may increasingly help draft consultation notes, summarise patient histories and prepare discharge documentation. However, systems can omit symptoms, misinterpret medication names or include information that was never discussed.

What to learn: Electronic health records, review of AI-generated notes, data privacy, documentation standards and safe clinical workflow design.

Change 3: Imaging-intensive specialities may evolve

AI can support selected tasks in radiology, pathology and other fields involving images or pattern recognition. Possible applications include prioritising cases, measuring findings and assisting with image interpretation.

These specialities involve more than pattern detection. Doctors must interpret findings in context, consider alternative explanations, communicate results and determine appropriate next steps.

What to learn: Speciality-specific AI tools, quality assurance, diagnostic validation, digital imaging workflows and multidisciplinary communication.

Change 4: Predictive and personalised healthcare may expand

AI may help identify patterns in electronic records, laboratory results, imaging, wearable-device data and remote monitoring. These applications may support risk assessment and follow-up planning.

A prediction is not a certainty. Poorly calibrated systems or data that do not represent the intended patient population can produce misleading estimates.

What to learn: Medical statistics, probability, sensitivity and specificity, risk prediction, remote monitoring and responsible data use.

Change 5: Human communication may become even more important

Patients need more than information. They need clear explanations, shared decision-making and care that considers their preferences, family circumstances and concerns.

AI may help prepare information, but doctors must still understand patients’ needs, communicate uncertainty and support informed decisions.

What to develop: Empathy, active listening, explaining complex medical information, informed consent, breaking difficult news and cultural awareness.

Change 6: Doctors may help evaluate healthcare technology

Healthcare organisations need clinical expertise when assessing whether AI systems are safe, useful and compatible with existing workflows. Some doctors may pursue opportunities in clinical AI evaluation, digital health implementation, research, quality improvement or medical technology development.

These are possible career directions, not guaranteed new jobs in every healthcare market.

What to learn: Clinical research, health informatics, evidence appraisal, AI evaluation, medical ethics and collaboration with technical teams.

3. Which Medical Tasks May Be Automated?

It is more useful to examine individual tasks than to label an entire medical speciality as safe or unsafe.

This is a qualitative overview, not a measured forecast of automation rates.

4. Will AI Replace Medical Doctors by 2036?

There is no reliable basis for claiming that AI will replace all medical doctors within the next 10 years.

A more defensible expectation is that many doctors will work alongside increasingly capable AI systems, with the degree of change varying across clinical roles.

Some tasks may become faster or require less manual work. Other tasks may require doctors to review automated outputs, investigate unusual findings, communicate decisions and manage cases where information is incomplete.

The effect on employment will depend on more than technical capability. Healthcare organisations must consider patient safety, clinical validation, privacy, regulation, costs, responsibility for errors and whether a tool works for its intended population.

The World Health Organization has highlighted the importance of ethics, oversight and governance in healthcare AI.

Read the WHO guidance on ethics and governance of AI for health →

5. How Could AI Affect Different Medical Specialities?

Radiology

AI may assist with selected image-analysis tasks, measurements, prioritisation and identification of abnormalities. Radiologists will still need to interpret findings in clinical context, evaluate uncertainty and communicate results.

Pathology

Digital pathology and image-analysis tools may support selected tasks involving tissue samples. Pathologists will still need to assess the specimen, integrate clinical information and interpret complex or uncertain findings.

General practice and internal medicine

AI may help summarise patient histories, review selected clinical information and support documentation or decision-making. Doctors will still need to manage multiple conditions, assess the whole patient and make decisions based on individual circumstances.

Surgery

AI and robotic systems may support planning, navigation and selected procedural tasks. These technologies do not remove the need for surgical training, technical judgement, patient selection or the ability to respond to unexpected events.

Psychiatry and mental healthcare

Digital tools may assist with selected administrative or monitoring tasks. Sensitive clinical conversations, risk assessment, therapeutic relationships and individualised care remain important. Automated tools must not be assumed to provide appropriate care in every situation.

Emergency and critical care

AI may help analyse patient-monitoring data and flag possible deterioration. Doctors must still assess the patient, interpret alerts, prioritise interventions and respond to changing clinical conditions.

Important: This comparison describes possible areas of technological influence. It does not rank specialities by job security or predict which speciality will experience the most employment disruption.

6. Seven Skills Doctors Should Develop

1. AI literacy for healthcare

Understand AI, machine learning, generative AI and clinical decision-support systems. Learn what these systems can do, where they fail and how their performance is evaluated.

• What problem is the tool designed to solve?
• Has it been evaluated in a population similar to your patients?
• What are its known limitations?
• Who reviews its output and remains accountable for the clinical decision?

2. Data literacy and medical statistics

Learn to interpret diagnostic accuracy, confidence intervals, sensitivity, specificity, predictive values and the difference between correlation and causation. These concepts help doctors assess whether a tool is clinically useful.

3. Safe use of generative AI

Learn to use approved tools for literature summaries, educational material, draft documentation and workflow support. Verify important claims against reliable clinical sources. Never enter identifiable patient information into consumer AI tools unless the tool and use are explicitly approved under applicable privacy, security and institutional policies.

4. Digital health and clinical informatics

Become comfortable with electronic health records, digital prescribing, telemedicine, remote monitoring and clinical information systems relevant to your workplace. Understand how poor-quality data can produce unreliable results.

5. Evidence appraisal and research skills

Learn to read clinical studies, assess methodology, recognise bias and understand whether results are clinically meaningful. Doctors need to distinguish validated tools from products supported by limited evidence.

6. Communication, ethics and patient trust

Develop skills in informed consent, shared decision-making, explaining uncertainty and communicating with patients who may be using AI tools themselves. Follow applicable professional and institutional requirements concerning AI use.

7. Speciality-specific technology skills

Prioritise technologies relevant to your field. Examples include imaging AI in radiology, digital pathology, AI-supported monitoring in intensive care, clinical decision support in primary care and digital planning or robotic assistance in selected surgical settings.

You do not need to study every emerging technology at once. Start with tools that address real problems in your clinical environment.

7. AI Courses Doctors Can Consider

The following courses and learning resources are offered by established medical, academic or healthcare organisations. They cover different levels, from short introductions to more detailed programmes. Check current availability, fees, schedules, eligibility and course content on the provider’s official website before enrolling.

Which course should you choose first?

• Medical student or intern: Start with the AMA AI in Health Care Series or Harvard’s short HMX course.
• Practising general practitioner: Begin with an introductory course, then focus on safe AI use, documentation, clinical decision support and evidence appraisal.
• Medical specialist: Build general AI literacy and then investigate validated tools and training relevant to your speciality.
• Clinical leader or researcher: Consider a more detailed specialisation or advanced clinical programme, alongside training in governance, research methods and implementation.
• Doctor interested in healthcare technology: Consider further study in statistics, health informatics, data science and machine learning.

You do not need to complete every course. Choose one suitable starting point, apply what you learn and add further training when it addresses a genuine skill gap.

8. A Practical 12-Month Learning Plan for Doctors

You can begin preparing without leaving your job or immediately committing to another degree.

9. What Should Medical Students Learn Now?

If you are considering medicine as a career, do not treat AI knowledge as a substitute for medical education.

1. Build a strong foundation in biology, anatomy, physiology and other sciences required for medical education.
2. Develop disciplined clinical reasoning and learn to question evidence.
3. Build communication skills and the ability to work with patients and multidisciplinary teams.
4. Become comfortable with digital tools and learn the fundamentals of AI.
5. Understand medical ethics, confidentiality and patient safety.
6. Continue learning throughout medical school, internship, residency and professional practice.

Do not choose a speciality solely on the assumption that AI will eliminate or protect particular jobs. Speciality choice should also consider your interests, aptitude, working style, training opportunities and the healthcare needs of the population you want to serve.

10. Should Doctors Learn Programming?

Not every doctor needs to learn Python, build machine-learning models or become a software developer.

For most clinicians, the first priority is understanding AI well enough to use approved tools safely, interpret their limitations and evaluate their clinical value.

Programming and deeper technical training can be useful if you want to pursue medical research, clinical AI development, health data science, digital health leadership or a role that combines medicine with technology.

A possible technical learning path

1. Basic statistics and data analysis.
2. Introductory programming, such as Python.
3. Fundamentals of machine learning.
4. Clinical data management and research methods.
5. Model evaluation, bias and responsible AI.
6. A supervised project addressing a genuine healthcare problem.

Technical skills should complement, not replace, clinical knowledge. Healthcare AI development requires attention to patient safety, data quality, validation and applicable legal and regulatory requirements.

11. Mistakes Doctors Should Avoid When Adopting AI

Trusting AI-generated answers without verification

AI can produce convincing but incorrect statements. Verify important clinical information using trusted evidence and professional guidance.

Entering confidential patient information into unapproved tools

Patient information must be handled in accordance with applicable privacy laws, professional duties and institutional policies. Use only tools and workflows approved for the intended data.

Allowing AI to weaken clinical fundamentals

Over-reliance on automated recommendations may undermine independent reasoning. Continue practising history-taking, examination, diagnostic reasoning and other core clinical skills.

Choosing a course only because of its brand name

Compare the syllabus, learning outcomes, level, duration, cost and practical relevance. A short course that addresses a real skill gap may be more useful than an expensive programme that does not match your needs.

Assuming every AI product is clinically validated

A demonstration, research paper or commercial claim is not equivalent to proof that a tool is safe and effective for a particular clinical use. Check available evidence and relevant regulatory status.

Treating AI adoption as a competition between doctors and machines

The objective should be safe, effective patient care. Doctors should understand both the capabilities of AI and the situations in which human judgement, further investigation or a different approach is required.

12. How Can Doctors Remain Valuable in an AI-Enabled Healthcare System?

A practical strategy is to combine four areas of professional development.

Clinical excellence

Maintain strong speciality knowledge, sound judgement and the ability to manage complex or uncertain cases.

Technology literacy

Understand relevant AI tools, digital records, clinical decision support and data privacy.

Human-centred care

Communicate well, respect patient preferences and make decisions collaboratively.

Continuous learning

Regularly review new clinical evidence, professional guidance and technology developments relevant to your work.

Doctors who combine these capabilities may be better prepared to adapt as their workplaces change. This is a strategy for improving professional resilience, not a guarantee of employment.

13. Frequently Asked Questions

Will AI replace doctors by 2036?

There is no reliable evidence that AI will replace all doctors by 2036. AI is likely to automate or assist with selected tasks, but its impact on whole jobs will depend on clinical performance, regulation, costs, workflows and patient acceptance.

Which doctors are most at risk from AI?

It is not possible to produce a reliable universal ranking of medical specialities by future job-loss risk. Some tasks involving standardised data analysis or repetitive documentation may be easier to automate than work requiring complex judgement, physical procedures or nuanced patient communication. Even within the same speciality, the impact can differ significantly.

Should I avoid becoming a doctor because of AI?

AI alone is not a sound reason to abandon an interest in medicine. Consider the training required, your interests, the work involved and the needs of the healthcare system. Be prepared to develop both clinical expertise and technology literacy.

Do doctors need to learn coding?

Most doctors do not need advanced coding skills simply to use AI responsibly in clinical practice. Programming becomes more relevant if you want to build healthcare technology, conduct technical research or work in health data science.

Are AI certificates compulsory for doctors?

There is no universal requirement for every doctor to hold an AI certificate. Requirements vary by employer, jurisdiction, speciality and role. Follow applicable professional and institutional requirements, and choose courses based on practical learning needs.

Can doctors use ChatGPT or other general AI tools for patient care?

Permitted uses depend on the tool, intended purpose, data involved, professional rules and local law. Never assume a general-purpose chatbot is an approved clinical decision-support system. Verify outputs, protect patient information and follow institutional policies.

How often should doctors update their AI knowledge?

There is no universal interval that applies to every doctor. Review developments regularly, follow speciality-specific guidance and seek further training when the tools, risks or responsibilities relevant to your practice change.

Conclusion: The Future Doctor Will Need Both Medical Expertise and Technology Literacy

AI is likely to change many aspects of medicine over the next decade. Some tasks will become faster or more automated, new digital workflows will emerge, and doctors will need to evaluate an increasing range of technology-assisted recommendations.

But the future of medicine cannot be understood by asking only whether AI can perform a task. It is also necessary to ask whether it can perform that task safely, consistently and appropriately for a particular patient—and who is responsible for the final clinical decision.

For medical students, the priority is to build strong clinical foundations while developing an understanding of technology. For practising doctors, the priority is to keep clinical skills current, learn to evaluate AI critically and choose training that addresses genuine needs in their field.

Research and Further Reading

1. US Food and Drug Administration: AI-Enabled Medical Devices
2. American Medical Association: Physician AI Use Survey
3. WHO: Ethics and Governance of Artificial Intelligence for Health
4. WHO: Guidance on Large Multi-Modal Models
5. US Office of the National Coordinator for Health Information Technology: Clinical Decision Support
6. AMA: Artificial Intelligence in Health Care Series
7. Harvard Medical School: AI in Medicine
8. Stanford Online / Coursera: AI in Healthcare Specialization
9. Harvard Medical School: AI in Clinical Medicine

Disclaimer: This article is for career education and general information. It is not a prediction of individual employment outcomes, a substitute for medical training or professional advice, or an endorsement of any particular AI product. Course information, fees and availability should be reconfirmed with the provider before enrolment.