The short answer: the jobs least likely to be automated combine physical work in unpredictable environments, hands-on care, complex human judgment, and trust-based relationships: nursing and therapy roles, skilled trades, teaching, counseling, and management. A 2025 study by the International Labour Organization and Poland’s NASK found that one in four jobs worldwide is potentially exposed to generative AI, with clerical jobs the most exposed, but that transformation, not replacement, is the most likely outcome for most occupations. Many of the fastest-growing U.S. occupations to 2035 are in healthcare, clean energy and data, according to the Bureau of Labor Statistics.
What the research says about AI and jobs
The ILO–NASK index, published in May 2025, combined nearly 30,000 occupational tasks with expert review and AI-assisted scoring. Its key findings:
- 25% of global employment is in occupations potentially exposed to generative AI, rising to 34% in high-income countries.
- Clerical jobs face the highest exposure, because generative AI could in theory automate many of their tasks.
- Some highly digitized cognitive jobs in media, software and finance also have growing exposure.
- Full job automation remains limited, because many tasks still require human involvement even when they’re done more efficiently.
- Exposure is not the same as job loss: the figures show potential exposure, and real effects depend on technology, infrastructure, skills and policy.
What makes work hard to automate
- Physical dexterity in unpredictable settings: repairing equipment in a customer’s home, installing solar panels on different roofs.
- Hands-on care: nursing, therapy, personal care.
- Complex judgment with accountability: diagnosing, managing risk, leading people.
- Trust and relationships: counseling, teaching, sales of complex products.
- Creativity tied to human context: design, strategy, research that defines the question as well as the answer.
Even in these jobs, AI will change tasks. The safest strategy is to combine a human-centered role with the ability to use AI tools well.
The fastest-growing occupations, 2025–2035
From the Bureau of Labor Statistics’ Occupational Outlook Handbook (projected growth and May 2025 median pay):
| Occupation | Projected growth, 2025–35 | Median pay (2025) |
|---|---|---|
| Nurse practitioners | 41% | $132,300 |
| Solar photovoltaic installers | 37% | $53,140 |
| Data scientists | 35% | $120,230 |
| Wind turbine service technicians | 30% | $64,120 |
| Medical and health services managers | 24% | $123,860 |
| Physical therapist assistants | 23% | $68,380 |
| Psychiatric technicians | 22% | $45,130 |
| Computer and information research scientists | 22% | $140,300 |
| Occupational therapy assistants | 21% | $72,300 |
| Ophthalmic medical technicians | 21% | $45,570 |
| Physician assistants | 21% | $135,880 |
| Information security analysts | 21% | $129,180 |
Healthcare and clean energy dominate the list, driven by an aging population and energy investment, alongside data and security roles. Several of these (nursing, therapy, solar and wind technician work) also combine physical and human skills that are hard to automate.
Careers that combine growth and resistance to automation
| Field | Examples | Why they’re resilient |
|---|---|---|
| Healthcare | Nurse practitioners, physician assistants, therapists and assistants | Hands-on care, clinical judgment, trust |
| Skilled trades and energy | Electricians, wind and solar technicians, industrial mechanics | Physical work in varied environments |
| Education and counseling | Teachers, school counselors, mental health counselors | Relationships, motivation, judgment |
| Management | Health services managers, operations managers | Accountability, people leadership |
| Data and security | Data scientists, information security analysts | Growing demand; AI changes the tools, not the need for judgment |
How to interview for these roles
- Show human skills with evidence: stories about patient care, de-escalation, teaching, leading a team. See teamwork interview questions.
- Show you use AI well: one example of using an AI tool to save time, plus how you checked its output. See how to talk about AI tools in an interview.
- Show adaptability and learning: employers expect skills to keep changing. See adaptability interview questions.
- Prepare role-specific questions: see our guides for physician assistants, occupational therapists, data scientists and security engineers.
If your current job is highly exposed
- List your tasks and mark which ones AI tools could already do or speed up.
- Move toward the human parts of your work: client relationships, judgment calls, training others.
- Learn to use AI tools in your current role, and document the results.
- Look at adjacent roles that use your experience, for example from administrative work into operations, project coordination or customer success.
Frequently asked questions
Which jobs are most resistant to AI and automation?
How many jobs are exposed to generative AI?
Which jobs are most exposed to generative AI?
What are the fastest-growing jobs to 2035?
How can I make my career more resistant to AI?
Sources
- International Labour Organization, One in four jobs at risk of being transformed by GenAI, new ILO–NASK Global Index shows (20 May 2025). Read on 2 October 2026.
- U.S. Bureau of Labor Statistics, Fastest Growing Occupations, Occupational Outlook Handbook. Read on 2 October 2026.
