Which Jobs Are Safe from AI? A Data-Backed Guide (2026)

As someone who has spent years managing a technical, non-client-facing data team, I have found AI implementation hard to navigate in my career. 2025 especially was the year of hearing executives say, "we don’t need people, just use AI," as justification for shorter project turnaround times, for not replacing contracted team members, and for the layoffs that followed. To a team like mine that already felt like it was drowning, AI seemed like a sign the apocalypse was near for all of our jobs.

While I am not in the camp of believing that AI is going to replace the entire workforce (there is a large gap between what executive teams believe, or are told, that AI can do and what it can actually do at this point), I do think it will continue to be used as a justification for layoffs without making the company look like it is in financial trouble (even if that might be what is really going on).

If you are a new grad trying to break into your first role, or someone who has lost their job because a company tried to have AI replace your work, it might feel particularly dire right now. You may be wondering if you need to make an entire career pivot, or if the degree you just got was worth the cost.

The good news is that the research tells a more nuanced story than the fear-driven headlines suggest. AI is reshaping the labor market (rapidly in some corners, and barely at all in others), and knowing where your skills sit on that map is one of the most useful things you can do while planning your comeback.

This guide pulls together the most credible data available (from the World Economic Forum, McKinsey, Goldman Sachs, the International Monetary Fund, and AI labs themselves) to show which jobs and industries face the highest automation risk, which are the safest, and what actually separates the two

The Big Picture: Alarming Numbers That Cut Both Ways

Before looking at specific roles, it helps to understand the scale. The headline figures sound frightening until you notice they point in two directions at once:

●        World Economic Forum (Future of Jobs 2025): 92 million jobs displaced by 2030, but 170 million new ones created, for a net gain of 78 million jobs.

●        McKinsey Global Institute: up to 30% of hours worked in the US economy could be automated by 2030.

●        Goldman Sachs (2023): AI could automate tasks equivalent to 300 million full-time jobs globally, though most workers are expected to see their roles augmented rather than eliminated.

●        International Monetary Fund: roughly 40% of jobs worldwide will be affected by AI in some form.

●        WEF also found that 39% of the average worker’s skill set is expected to be transformed or become outdated between 2025 and 2030.

The word that matters most

Notice that most of this research measures tasks, not whole jobs. AI is very good at automating specific tasks (data entry, drafting, summarizing), but far weaker at replacing an entire role built on judgment, physical presence, or human relationships. That distinction is the key to understanding your own risk.

Jobs and Industries Most at Risk

The research is remarkably consistent about where the pressure is greatest. It clusters around office, administrative, and data-processing work, the kind of predictable, screen-based tasks that generative AI handles well.

An analysis of 784 US occupations (using the Wharton Budget Model) found that the 50 most exposed jobs average 86.3% task exposure, nearly 2.9 times the national average of 29.84%. Telemarketers top the entire list at 96.25% exposure, meaning almost every task in the role could be handled by current or near-future AI.

Highest-Exposure Job Titles (share of tasks AI could perform)

Job Title	Task Exposure
Telemarketers	96%
Data entry clerks	Very high
Customer service representatives	Very high
Administrative assistants / secretaries	Very high
Bookkeeping and accounting clerks	High
Bank tellers	High
Postal service clerks	High
Graphic designers (routine production)	Rising

The WEF names cashiers, ticket clerks, administrative assistants, postal clerks, bank tellers, and data entry clerks among the fastest-declining roles in absolute numbers. Graphic designers appeared on that declining list for the first time in 2025, a direct result of generative AI’s ability to produce visual content.

Highest-Exposure Industries

●        Technology and IT: about 92% of IT roles are expected to be transformed by AI, with mid-level (40%) and entry-level (37%) positions hit hardest.

●        Retail: roughly 65% of tasks face automation risk from AI and robotics by the end of 2025.

●        Financial services: routine processing, compliance, and junior analyst roles are being restructured at scale, even as demand grows for data scientists and fintech specialists.

Jobs and Industries Least at Risk

Now the more encouraging half of the picture. The same 784-occupation analysis identified 50 occupations with 0.0% AI exposure (just 6.4% of all jobs studied). Construction and extraction trades made up 33 of those 50 safest roles, roughly two-thirds of the entire zero-exposure group.

AI labs see the same pattern in real usage. Anthropic’s analysis of millions of actual AI interactions found that about 30% of workers fall into a "zero exposure" group whose day-to-day tasks rarely show up in AI use at all. The examples are telling: cooks, bartenders, lifeguards, motorcycle mechanics, and dishwashers, all physical, in-person work that a chatbot cannot deliver through a screen.

On the higher-paid end, a separate AI-Resistant Careers Index (built on US Department of Labor O*NET data) ranked nurse anesthetists as the single most insulated role, pairing a top resistance score with a median salary of about $195,000.

Lowest-Risk Job Titles

Job Title	AI Exposure
Nurse anesthetists / nurse practitioners	Very low
Physical and occupational therapists	Very low
Mental health counselors	Very low
Electricians, plumbers, HVAC technicians	0%
Construction and extraction trades	0%
Physician assistants	Very low
Cooks, bartenders, personal care aides	Very low
Choreographers and skilled creatives	Very low

Healthcare demand is a bright spot: nurse practitioner roles are projected to grow more than 45% by 2032, far faster than the average occupation, as AI augments rather than replaces clinical work.

Lowest-Risk Industries

●        Skilled physical trades (electrical, plumbing, HVAC, construction)

●        Healthcare and caregiving (nurses, therapists, aides, physician assistants)

●        Personal and in-person services (food preparation, hospitality, personal care)

●        Education (teachers, instructors, and school leadership)

●        Senior creative direction and strategy (original ideas and accountability, not routine production)

What Actually Makes a Job AI-Resistant

This may be counterintuitive: automation risk does not track neatly with education, income, or prestige. A historian with a doctorate can face higher exposure than a phlebotomist with a certificate. A six-figure data analyst can be more vulnerable than a nursing assistant earning a fraction of that. What protects a role is not credentials, it is the presence of skills machines still struggle to replicate:

●        Physical presence and manual dexterity in unpredictable, real-world environments

●        Emotional intelligence and interpersonal trust (care, persuasion, negotiation)

●        Real-time judgment and accountability for high-stakes outcomes

●        Original creativity and strategy, as opposed to routine content generation

A white-collar title and a high salary are not signals of safety. In fact, some office roles are more exposed than hands-on jobs that pay less. The four traits above are the better gauge of how secure your specific role really is.

If You Have Been Laid Off, Here Is What to Do With This

You do not need to abandon your field or retrain from scratch. You need to reposition. Five practical moves:

1. Audit your tasks, not your title. Break your last role into its daily tasks. The routine, repeatable ones are the exposed ones. The judgment-heavy and people-heavy ones are your moat.

2. Move toward the human edge. Within almost any field there are roles that lean more on relationships, strategy, and accountability. Steer toward those.

3. Learn to work with AI, not against it. WEF lists analytical thinking, resilience, and AI literacy among the most in-demand skills through 2030. Being the person who uses AI well is far safer than competing with it.

4. Consider adjacent, lower-risk roles. Skills often transfer sideways. Customer service experience can move into care coordination or client success, where the human element is the point.

5. Watch the entry-level squeeze. Anthropic flagged a 6% to 16% drop in employment among workers aged 22 to 25 in highly exposed occupations, driven mainly by slower hiring rather than layoffs. If you are early-career, prioritizing an AI-resilient field early pays off.

What This Actually Means for You

No job is completely immune, and no job is disappearing overnight. AI is automating tasks far faster than it is eliminating whole careers, and the roles that combine human judgment with physical or emotional presence remain the most durable. If you are between jobs right now, the smartest move is not to panic about automation. It is to position yourself where human skills are becoming more valuable, not less.

Sources

World Economic Forum, Future of Jobs Report 2025

McKinsey Global Institute, workforce automation projections (2024 to 2030)

Goldman Sachs, generative AI labor market research (2023)

International Monetary Fund, AI and the global workforce

Wharton Budget Model AI exposure dataset (via EDsmart, 2025)

Anthropic Economic Index, real-world AI usage analysis (2025)

Resume Now, AI-Resistant Careers Index (O*NET data)

US Bureau of Labor Statistics occupational growth projections

About the Author

Corporate Kate has spent nearly 15 years inside corporate tech, managing large teams and making the hiring decisions most job seekers never get to see. She holds a bachelor’s degree in Finance and writes about layoffs, careers, and money.