A medical transcriptionist in Phoenix lost her contract in January. Not to a cheaper vendor. Not to outsourcing. To a single API call that her hospital now runs in real time during appointments. She had 14 years of experience. Her job scored a 10 on the AI Displacement Index. Her outlook: -8%.
Meanwhile, a software engineer at a mid-size fintech is using AI agents to ship code 3x faster. His job also scores 8-9 on exposure. His job outlook: +25%.
Same high score. Completely different futures. That is the thing most coverage on AI agents replacing workers gets wrong.
The real exposure picture
42% of US jobs score 7 or higher on AI exposure. That is 59.9 million jobs and $3.7 trillion in wages sitting in the restructuring zone right now.
The question is not "will AI agents change my job?" They already are. The question is whether your role is being amplified or eliminated.
What Most People Get Wrong About AI Agents
People hear "AI agents in the workplace" and picture robots taking desks. The reality is quieter and faster. Autonomous AI in the workplace does not announce itself. It just handles the task before you get to it.
The popular assumption: low-skill, low-wage jobs go first. The data says the opposite.
Jobs paying under $35K average a 3.4 AI exposure score. Jobs paying over $100K average 6.7. Bachelor's degree holders average 6.7 exposure. No degree: 4.1.
Your degree does not protect you. For knowledge workers, it amplifies your exposure. The more cognitive the role, the more surface area for autonomous AI to claim.
The plumber is not threatened. The paralegal is. That inversion is the story. And most people with white-collar jobs are not looking at their exposure score yet.
The Danger Zone: Score Plus Outlook
A high AI exposure score is not a death sentence. It depends on what you pair it with.
Here is how to actually read the risk timeline. Score alone is not enough. The job outlook is the other axis.
-
Score 9-10 + declining outlook. This is disruption happening now. Medical transcriptionists. Certain data entry clerks. Telecom equipment operators. The contract is already gone.
-
Score 7-8 + flat or declining outlook. Two to three year restructuring window. The role will still exist but it will look different. The people who adapt early keep their seats.
-
Score 7-9 + strong growth outlook. The paradox zone. High exposure but expanding demand. Software developers live here. So do certain financial analysts who can use AI to do the work of three people.
-
Score 0-3 + any outlook. Physical, relational, unpredictable. Electricians score 1. Nurses score 2. Physical therapists score 3. No autonomous AI can do what they do in the field.
Only 3% of occupations score 9-10. The bulk of exposure, 42% of all jobs, sits in the 7-8 range. Not eliminated. Restructured. That distinction matters enormously for how you plan.
Timeline by score band
Score 9-10: disruption is current. Score 7-8: 2-3 year restructuring window. Score 5-6: 5+ years before major change hits. Know your band.
Same Industry, Different Futures
Healthcare is the clearest illustration of how exposure works at the task level, not the job level.
Surgeons score 3. Radiologists score 7. Both are physicians. Both spent a decade in training. But one interprets images, the other operates in the physical world with tactile judgment and real-time adaptation. Autonomous AI can match pattern recognition. It cannot hold a scalpel.
81% of physicians now use AI daily, up from 38% in 2023. The transition happened faster than anyone predicted. But the radiologist who uses AI to read 400 scans a day is not being replaced. She is being out-competed by radiologists who use AI, versus radiologists who do not.
The question is not "will AI agents take my job?" It is "will someone using AI agents take my job?"
The same dynamic plays out in sales. VP of Sales scores 6. Manageable. But the SDRs reporting to that VP score 8. The second-order effect hits the team before it hits the manager. The VP who does not see that coming loses half their headcount budget and wonders why pipeline dried up.
Where does your role land?
500+ occupations scored 0-10. Free. Takes 60 seconds.
What Actually Protects You
Andrej Karpathy published a 342-occupation analysis on March 15, 2026. The most durable jobs share a specific profile. Not prestige. Not salary. A specific type of task composition.
Roles that are hard to automate tend to combine physical presence, unpredictable environments, and real-time human judgment. Electricians score 1. HVAC technicians score near 0. Plumbers score 1. That is why 42% of Gen Z is now pursuing trades. They read the data before the white-collar world did.
But for knowledge workers already mid-career, pivoting to trades is not the answer. The answer is positioning inside your current role.
AI skills command a 56% salary premium right now. The gap between who adapts and who does not is already showing up in compensation data.
Three things you can act on immediately:
Score your specific tasks, not your title. "Software developer" and "medical transcriptionist" are both high-exposure job titles but they sit on opposite ends of the outlook axis. Pull up the task list for your role on O*NET and run it against what AI agents can currently do. You will find your real exposure level is more specific than any headline number.
Find the adjacent task that AI cannot do. In every high-exposure role there are tasks that require external judgment, stakeholder trust, or physical context. The software developer who handles client architecture decisions is safer than the one who only writes boilerplate. Move toward those tasks now, before the restructuring forces you to.
Build AI operator skills in your specific domain. The 56% salary premium is not going to people who know AI generally. It is going to people who know how to deploy AI agents in a specific context. A nurse who uses AI to triage notes is not threatened by AI. She is the person who trains the next system.
The adaptation premium
Workers with documented AI skills command a 56% salary premium over peers in the same role without them. This gap will widen as autonomous AI becomes standard infrastructure.
Bottom Line
The conversation about autonomous AI in the workplace is stuck on the wrong question. "Which jobs will AI take?" is less useful than "which tasks inside my job are already gone?"
The global average exposure score is 5.3 out of 10. That means most jobs are not in immediate danger. But that number hides enormous variance. A radiologist and a surgeon both appear in "healthcare." One has a 7. One has a 3. The average tells you nothing.
The data points that actually matter: your score, your job outlook, and how your task composition maps to what agents can do today versus what they cannot do yet. That intersection is your real risk profile. Not the headline number.
The full breakdown covers 500+ occupations across all major sectors, scored on the same 0-10 scale and cross-referenced against employment outlook, wage trajectory, and task-level exposure. If you are making a career decision in the next 12 months, that is the analysis you want to run first.
The workers who will look back on this period without regret are not the ones who panicked. They are the ones who looked at specific data, understood their actual position, and moved with purpose before the window closed.
In a restructuring, speed of accurate information beats everything else.
Find out where you stand
500+ occupations scored 0-10 on AI displacement risk. Free.
Keep Reading
Methodology: AI Displacement Scores are calculated using the JobHunter AI Displacement Index, which analyzes 500+ occupations across 12 risk factors including task automation potential, historical automation patterns, AI capability trajectories, and labor market dynamics. Data sources include Stanford's AI Index Report, Anthropic's capability research, Bureau of Labor Statistics employment projections, and O*NET task databases. Scores are updated quarterly. Learn more about our methodology.
Related AI Displacement Scores: Customer Service Representatives · Secretaries And Administrative Assistants · Bookkeeping Accounting And Auditing Clerks