A radiologist in Boston just lost a contract renewal to an algorithm. Not because she was bad at her job. Because the algorithm read 47,000 scans last quarter without taking a day off, and the hospital CFO noticed.
She still has work. For now. But her score on the AI Displacement Index is 7. Which means the restructuring has already started. She just doesn't feel it yet.
That gap between the data and the feeling is the most dangerous place to be in 2026.
We scored 500+ occupations on a 0-10 scale measuring AI exposure. Not hype. Not speculation. Actual task-level analysis of what AI can do today versus what each role requires. Here are 50 numbers that change how you think about where you stand.
What You Think Protects You. Doesn't.
Most professionals run the same mental model: high salary means safe job. Advanced degree means leverage. Specialized knowledge means job security. The AI employment data for 2026 breaks all three assumptions at once.
Jobs paying $100K+ average a 6.7 AI exposure score. Jobs paying under $35K average 3.4. You read that right. Higher pay correlates with higher AI pressure. Not lower.
The salary paradox
$100K+ roles average 6.7 exposure. Under $35K roles average 3.4. The plumber is safer than the product manager. That's not a fluke. That's physics.
Bachelor's degree holders average 6.7 AI exposure. No degree: 4.1. The credential that was supposed to future-proof you is now a flag for the roles AI is targeting hardest.
Why? Because high-pay, high-credential jobs are built around information processing, pattern recognition, and structured decision-making. Which is exactly what modern AI does well. The plumber installs pipe in a 100-year-old house with a weird layout. No model handles that yet.
42% of Gen Z are now pursuing trades. Plumbers and HVAC technicians score 0-2 on the displacement index. That's not anti-intellectualism. That's actuarial thinking.
The Number That Matters More Than Your Job Title
The global average AI exposure score across all 500+ occupations we analyzed: 5.3 out of 10. That sounds manageable. Until you look at distribution.
42% of US jobs score 7 or higher. That's 59.9 million jobs. $3.7 trillion in annual wages. These aren't jobs that AI will eliminate cleanly. They're jobs getting restructured task by task, quarter by quarter, in ways that are hard to see until they're already done.
Only 3% score 9-10. Near-full automation. The bulk of the danger zone is 7-8. That's the restructuring band. Jobs that still exist but look fundamentally different in 36 months.
Score 9-10 means disruption now. Score 7-8 means 2-3 years. Score 5-6 means 5 or more. The timeline is embedded in the number. Most people don't know their number.
Andrej Karpathy published a 342-occupation analysis on March 15, 2026, and the pattern held across every sector he examined. Title is a poor proxy for risk. Tasks are the real unit of analysis.
Your job title is a container. What's inside determines your actual exposure. A "marketing manager" who spends 70% of their time on content briefs and reporting scores very differently from one who runs brand strategy and manages agency relationships. Same title. Completely different futures.
Same Industry. Three Different Futures.
Healthcare is the cleanest case study for what the AI workforce statistics actually mean in practice. Look at three roles inside the same hospital system.
-
Nurse: score 2. Physical assessment, patient relationship, real-time judgment in uncontrolled environments. AI can assist. It cannot substitute. Employment outlook: strong.
-
Radiologist: score 7. Pattern recognition in standardized image data. 81% of physicians now use AI daily, up from 38% in 2023. The Boston radiologist at the top of this article is the pattern, not the exception.
-
Medical transcriptionist: score 10, outlook -8%. This is what the danger zone looks like. The task is entirely language-based, structured, and repetitive. There is no moat left.
Surgeons score 3. Fine motor skill, spatial reasoning, live tissue variability. Radiologists score 7. Same hospital. Same prestige. Completely different risk profile. The data does not care about your training or your salary.
The danger zone combination
Score 9-10 + declining job outlook = act now. Medical transcriptionist is the clearest current example. Score 10. Outlook -8%. That combination is not a forecast. It's a current event.
The Second-Order Effect Nobody Talks About
Software developers score 8-9 on AI exposure. GitHub Copilot, Claude, GPT-4 are writing meaningful production code. The task overlap is real and large. So the field is collapsing, right?
Job outlook: +25% growth. High score. Booming demand. That's the paradox the AI job market statistics keep throwing up.
When AI raises developer productivity, more software gets built. More software means more systems to architect, integrate, secure, and maintain. The demand surface expands faster than the productivity gain. At least for now.
But look one level down the org chart and the picture gets darker. The VP of Sales scores 6. The SDRs who report to them score 8. The VP's judgment, relationships, and strategy have friction that AI can't replicate yet. The SDR's outreach sequences, prospecting research, and email personalization? That's core AI territory.
Second-order effects. The role at the top of the hierarchy absorbs the tool. The roles underneath it absorb the displacement. When you see a job title, ask what happens to the people one level below it when AI scales what the senior person does.
The senior person adopts the tool. The junior person becomes redundant. AI doesn't eliminate the job title. It hollows out the entry points into the career.
Physical therapists score 3. Electricians score 1. These are not low-skill jobs. They are high-coordination, spatially complex, contextually variable roles that AI cannot perform in the physical world. The market is waking up to this. 42% of Gen Z are already making the calculation.
What's your score?
500+ occupations scored 0-10 on AI displacement. Free. Takes 60 seconds.
What Survives. And Why.
The AI employment data for 2026 points clearly at what stays human. Not because of sentiment. Because of task structure.
-
Physical world complexity. Electricians score 1. Plumbers score near 0. Every installation is a unique environment. Unstructured. Variable. Impossible to delegate to a model with no hands.
-
High-stakes human judgment. Nurses score 2. Physical therapists score 3. Real-time decisions with real consequences, in direct relationship with another person. Accountability and presence matter here.
-
AI skill premium. Workers who can direct and deploy AI tools are commanding a 56% salary premium over peers who cannot. The tool doesn't replace you. Your ability to use it is now the differentiator.
The skills premium
56% salary premium for workers with demonstrated AI skills. That number applies across sectors. The platform changes. The premium doesn't.
The 3% of jobs scoring 9-10 are already in disruption. Not forecasted disruption. Current disruption. Medical transcriptionists, data entry operators, basic paralegal research, routine customer support scripting. These are not future losses. They are happening now, quietly, in quarterly workforce planning meetings.
The 42% scoring 7+ are in a slower wave. Two to three years of restructuring that looks survivable until it isn't. That's the group that most needs to act. And the group most likely to wait.
Three Things You Can Do With This Data Today
The AI workforce statistics are descriptive. The response has to be prescriptive. Here's where to start.
Find your actual score. Not your job title's score. Your task breakdown score. If you spend 60% of your week on pattern recognition and structured reporting, you score higher than your job title suggests. Be honest. The free tool covers 500+ occupations and takes 60 seconds.
Identify which tasks inside your role are scoring 8+. Then either delegate them to AI tools now and move upstream, or develop irreplaceable skills around the tasks that score 2-3 in your field. The 56% salary premium goes to people who actively manage this transition, not wait for it.
Check the second-order effect on your role. If the senior version of your job uses AI to do more, what happens to the junior version? If you're in the junior version, the window for moving upstream is shorter than it feels. The data timeline is 2-3 years for the 7-8 band. That's not far.
The deeper survival playbook goes further. Specific pivot strategies by score range. Skill investment priorities by sector. How to reposition inside a high-exposure role before the restructuring makes the choice for you. That analysis is in the full report.
But step one is the score. You can't navigate without a position fix.
Bottom Line
The most dangerous number in the AI job market statistics isn't the global average of 5.3. It's the gap between your score and what you think your score is. That gap is where careers stall quietly, before anyone sounds an alarm.
High score and booming demand. Low score and physical world friction. The jobs surviving aren't lucky. They're structured differently. Understanding the structure of your own role is the only actionable response to data this specific.
The number doesn't lie. The delay does.
Find out where you stand
500+ occupations scored 0-10 on AI displacement risk. Free.