Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →For most workers, the best-supported answer is that AI is more likely to change parts of a job than eliminate the entire job—but no global finding can predict what will happen to your specific role. Generative AI can automate some tasks, help people do others, and leave work that depends on human judgment or action largely in human hands. The International Labour Organization (ILO) estimates that one in four workers worldwide are in occupations with some degree of generative-AI exposure; it says transformation is more likely than redundancy overall because most jobs still involve tasks requiring human input. The ILO’s 2025 update measures potential exposure, not a forecast that one in four workers will lose their jobs.
What does “AI exposure” actually mean?
Exposure means that some tasks within an occupation could potentially be affected by generative AI. It does not mean the technology is already being used to do them, that a whole job can be automated, or that an employer will cut staff. A role can include tasks AI may assist with alongside work that still requires a person to interpret context, take responsibility, interact with others, or act in the physical world.
The ILO’s 2025 exposure index estimates task-level potential using occupational information, expert input, and AI predictions. Its analysis draws on 29,753 occupational tasks and 52,558 data points covering 2,861 tasks. That granularity makes it more informative than treating an occupation as one indivisible unit, but it remains a modeled estimate—not a count of jobs already replaced. The ILO’s refined global index puts 3.3% of global employment in its highest exposure category.
What the current evidence says—and does not say
| Finding | What it measures | What it does not establish |
|---|---|---|
| One in four workers worldwide are in occupations with some degree of generative-AI exposure (ILO, 2025). | Global occupational exposure to generative AI under the ILO’s framework. | That one in four workers will lose their jobs. |
| 3.3% of global employment is in the ILO index’s highest exposure category (ILO, 2025). | The share of employment in the index’s top exposure gradient. | That these jobs have already been automated or will disappear. |
| About one-third of vacancies across ten OECD countries were in occupations classified as highly exposed to AI (OECD, 2024). | Online vacancies across ten countries, classified as highly exposed relative to the study’s exposure distribution. | That one-third of workers will lose jobs; this is not a job-loss projection. |
| Software developer employment in the United States was projected to grow 17.9% from 2023 to 2033 (BLS, 2025). | A U.S. occupational employment projection over that period. | That AI caused the expected growth, or that every software developer’s job is secure. |
These figures use different populations, definitions, and methods, so they should not be combined into one estimate. The OECD result concerns online vacancies and its own AI-exposure measure, while the ILO figures concern global employment and generative-AI exposure. The U.S. Bureau of Labor Statistics projection is about expected employment change, not a causal estimate of AI’s effect. Its analysis also cautions that exposure categories do not distinguish between automation and augmentation. The OECD analysis and the BLS explanation of AI in employment projections describe their respective measures and limits.
#1 Best Overall
Why an exposed job may still change rather than disappear
A job title bundles many activities together. AI might handle or speed up some recurring digital tasks while a worker continues to review the output, resolve exceptions, communicate with clients, make context-sensitive decisions, or remain accountable for the result. Whether that division of work leads to fewer positions, different responsibilities, or higher output depends on more than technical capability: employers must also decide whether and how to adopt the tools.
The ILO’s occupation explainer describes its method as an estimate of potential effects on occupations, followed by a separate step to assess possible employment effects. It also cautions that the technology is still evolving, making the future difficult to predict. Read the ILO’s explanation of how generative AI may affect occupations.
Rank #2
Adoption is a separate question from exposure. In an interview published in September 2025, ILO Senior Researcher Paweł Gmyrek said, “For the time being, we are still mostly discussing exposure to generative AI.” The interview reported that 9.4% of surveyed Polish workers said their employer had officially introduced generative-AI tools, based on a late-2024 survey. That is a specific result for surveyed workers in Poland, not a global workplace-adoption rate. Read the ILO interview with Paweł Gmyrek.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How to assess what might change in your own job
There is no reliable universal checklist that labels particular tasks “safe” or predicts whether your employer will eliminate a role. You can, however, make the question more concrete by examining your own work and workplace:
- Name your occupation and location. Findings for global employment, U.S. projections, and vacancies in ten OECD countries do not describe every country or labor market equally.
- List the recurring tasks that fill your working week. Separate the activities rather than treating your job title as a single unit.
- Note what kind of work each task involves. For example, distinguish routine digital work from activities involving physical presence, interpersonal judgment, accountability, or decisions that depend on specific context. These are useful distinctions for discussion, not guarantees that a task or job is protected.
- Find out whether your employer has introduced AI tools. A task’s potential exposure does not tell you whether your workplace has deployed a tool or changed staffing.
- Ask what the tool is expected to do and who checks the result. The practical change may be assistance, task automation, a new review duty, or some combination; the exposure label alone cannot determine which.
This exercise is a way to frame questions about your work, not a validated personal risk score. An exposure figure by itself is not enough reason to quit, change careers, or buy a training program.
Quick Recap
Best Value
What you can reasonably conclude
- Generative AI may affect individual tasks without replacing the whole occupation.
- Exposure measures describe modeled potential, not observed job losses or certain outcomes for individual workers.
- Whether a task is technically exposed and whether an employer adopts a tool are separate questions.
- Occupation, location, workplace decisions, and the mix of tasks in your role all matter to a personal assessment.
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.
