For most people the evidence points to change before replacement. AI is more likely to take over or reshape some of your tasks than to remove your whole occupation. That is a pattern across many jobs, though, not a forecast for yours. Whether your own work is at risk depends on which tasks fill your week, whether your employer actually uses the tools, and where you work. This article explains what the main studies measure and then gives you a way to check your own role.
What the strongest evidence says
The International Labour Organization (ILO) published a research brief on 20 May 2025, Generative AI and jobs: A 2025 update. It states: “One in four workers across the world are in an occupation with some degree of GenAI exposure, but because of the continued need for human input, most jobs will be transformed rather than made redundant.”
As an Amazon Associate I earn from qualifying purchases.
The ILO also says plainly that prediction has limits. In its explainer How might generative AI impact different occupations? (20 May 2025), it writes: “While it is not possible to predict the future – particularly as the technology is still evolving – ILO researchers first developed a methodology in 2023, and later refined it in 2025, to estimate the potential effects of generative AI on existing occupations, and then in a second step, on employment.”
Recommended Free Tools
What “exposure” actually means
Exposure is a modeled estimate of how much of an occupation’s work generative AI could potentially do or help with. It is not a count of jobs already replaced, and it does not say what any employer will decide.
#1 Best Overall
The ILO’s 2025 working paper, Generative AI and Jobs: A Refined Global Index of Occupational Exposure, builds its index at the task level. It draws on 29,753 occupational tasks, with 52,558 data points covering 2,861 tasks. Scores combine human input, expert discussion and AI predictions. Task-level scoring is why the index is more granular than a label such as “office worker”. It still produces potential, not observed outcomes.
The numbers, with their limits
| Figure | Source and date | What it covers | What it does not say |
|---|---|---|---|
| 1 in 4 workers in an occupation with some GenAI exposure | ILO, May 2025 | Global employment, occupational exposure | Not the share of workers who will lose jobs |
| 3.3% of global employment in the highest exposure gradient | ILO working paper, May 2025 | Global employment; clerical work is the most exposed broad occupational group in the refined index | Not a count of displaced clerks |
| About one-third of vacancies in highly exposed occupations | OECD, 2024 | Online vacancies in ten OECD countries, using a particular AI exposure measure with high exposure defined relative to its distribution | Not a prediction of one-third job losses |
| 17.9% projected growth in software developer employment, 2023–33 | U.S. Bureau of Labor Statistics, March 2025 | U.S. employment projection | Not proof that AI causes the growth |
| 9.4% of surveyed workers said their employer had officially introduced GenAI tools | ILO interview with Paweł Gmyrek, 29 September 2025, citing a late-2024 survey | Polish workers only | Not a global or European adoption rate |
Do not blend these into one percentage. They measure different populations (workers, vacancies, projected jobs), use different definitions of exposure and cover different periods.
Why exposure is not the same as job loss
Jobs are bundles of tasks
Most occupations contain several kinds of work. AI may draft, summarize or classify one part while another part still needs a person to decide, be present or be accountable. That mix is the ILO’s reason for expecting transformation more often than redundancy.
Free tools Windows power users keep installed
One-click scans. No signup required.
Exposed occupations can still grow
Software development is widely discussed as AI-exposed, yet the BLS’s March 2025 article AI impacts in BLS employment projections reports a 17.9% projected increase in U.S. software developer employment from 2023 to 2033. The projection does not show that AI drives the growth. It does show that “exposed” cannot be read as “shrinking”.
Automation and augmentation look the same in a score
A high exposure score can mean a tool might do a task, or only that it might help someone do it faster. Those lead to very different outcomes for a worker, and a single score generally cannot tell them apart.
Capability is not adoption
Gmyrek said in the ILO interview that exposure reflects the potential to automate tasks, not the immediate automation of a whole occupation. “For the time being,” he said, “we are still mostly discussing exposure to generative AI.” A task can be technically exposed while your workplace has deployed nothing.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How to assess your own job
This is a set of questions to work through, not a validated prediction test. Nothing below guarantees safety or danger.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →- Name your occupation and location. Exposure data, projections and employer behavior differ by country and by occupation.
- List the recurring tasks in a typical week and roughly how much time each takes. Aim for 10 to 20 specific tasks rather than a job title.
- Mark each task as mainly digital and repeatable, or not. Drafting, data entry, standard summaries and routine correspondence are the kind of digital text and information work that generative AI is aimed at. Work that depends on physical presence, interpersonal judgment, accountability or context-specific decisions is less obviously a fit. This is a way to organize your thinking, and it does not rule anything in or out.
- For each digital task, ask whether a tool would replace the person or speed them up. If a tool would speed you up, your role may change in how it is done. If a tool could do the whole task unsupervised, that part of your workload is the part most at stake.
- Check what your employer has actually introduced. Ask whether tools are officially approved, what policies exist, and what they are meant for. Adoption is often slower than capability.
- Look up official projections for your occupation. In the U.S., the BLS Employment Projections are a starting point. Treat them as projections with uncertainty, covering a defined period.
A role where most weekly time falls into the digital, repeatable group deserves closer attention than one where it does not. Even then, the useful response is usually to learn how the tools handle your own tasks. An exposure label alone is not a reason to quit, change careers or pay for training.
Best Value
What is not yet known
The evidence describes potential exposure and projected employment. It is not a definitive tally of jobs already displaced by generative AI, and the technology is still changing. Nobody can responsibly give you a personal replacement probability from these studies alone. You can, however, find out which of your tasks are most likely to change first.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
