Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →No one can responsibly name a current list of jobs that artificial intelligence will certainly eliminate. The best global evidence measures how much a job’s tasks could be affected by generative AI, not how many jobs will disappear. The International Labour Organization’s 2025 assessment finds clerical work has the highest exposure, while some highly digitized media, software and finance occupations are also increasingly exposed. Its central finding is that transformation is more likely than complete replacement because most occupations still include work requiring people.
That distinction matters for anyone making a career or financial decision. A job title is a bundle of tasks, and exposure can change the mix of those tasks without ending the occupation.
What the evidence can—and cannot—say
The ILO estimates that one in four workers worldwide are in an occupation with some degree of generative-AI exposure. That is a measure of potential task impact, not a forecast that one in four jobs will be lost. Only 3.3% of global employment falls into the ILO’s highest exposure category, and that share differs by gender and by national-income level.
The ILO says it is not possible to predict the future while the technology is still evolving. Employer adoption, workplace rules, infrastructure, worker skills and customer preferences all influence whether an exposed task is automated, assisted or redesigned. A country-level employment forecast therefore requires local occupational and labour-market data, not a global exposure percentage.
Which occupational groups show the most exposure?
Clerical work
Clerical occupations remain the most exposed group in the ILO’s 2025 GenAI index. Many duties involve structured digital information—entering, sorting, summarising, classifying or responding to text—which current models can perform or support. Exposure still varies inside the group: a role with a few automatable transactions is different from one in which most tasks are routine and digital.
Media, software and finance-related work
The ILO also reports increasing exposure in some strongly digitised media-, software- and finance-related occupations as models improve at text, voice, image, video and other specialised capabilities. “Increasing exposure” means more tasks are technically within reach; it does not mean every journalist, developer, analyst or finance worker faces the same outcome.
| Pattern in the evidence | What it means | What it does not establish |
|---|---|---|
| Clerical occupations have the highest average GenAI exposure. | Digital, structured information tasks are often easier for models to process. | That every clerical job, employer or country will eliminate the role. |
| Some media, software and finance occupations show rising exposure. | New model capabilities can reach more specialised digital tasks. | A timetable for those occupations to disappear. |
| Exposure differs among tasks inside one occupation. | Workers may see particular duties change while the job remains. | That an occupational average predicts an individual worker’s outcome. |
Exposure is not the same as automation risk
“Exposure” asks whether an AI system could affect a task. “Automation risk” asks whether an employer is likely to hand that task, or an entire workflow, to a system. Those are different questions.
OECD analysis of OECD countries places IT professionals, business professionals, managers and chief executives, and science and engineering professionals among the occupations most exposed to AI capabilities. These jobs can use AI extensively while retaining responsibilities that require non-routine judgement, accountability, social interaction, creativity or context. Conversely, work that appears less exposed to generative AI can still be affected by robotics or other automation technologies.
| Measure | Question it answers | What you should not infer |
|---|---|---|
| AI capability exposure | Could current or near-term systems perform or assist parts of the work? | That employers will adopt them, or that headcount will fall. |
| Observed adoption | Are organisations actually deploying systems in this workflow? | That adoption will have the same effect in every workplace. |
| Hiring or vacancy change | Are skill requirements or job postings changing? | That a short-term change proves permanent job destruction. |
| Employment displacement | Did people lose jobs or hours after a change? | That a modelled exposure score is a layoff count. |
How the ILO’s 2025 index was built
The ILO combined task-level information, worker input, expert validation and AI-assisted predictions. It began with a representative Polish occupational classification containing 29,753 tasks and collected 52,558 assessments of perceived automation potential covering 2,861 tasks. The task predictions were then extended to ISCO-08 occupations, producing a global assessment of 436 detailed occupations and applying the estimates to labour-force survey data from more than 140 countries.
The index uses four exposure gradients. They reflect both an occupation’s average task exposure and how much its tasks vary. A high, consistent score means many tasks are exposed; a similar average created by a few highly exposed tasks and many unaffected tasks implies a different transition. Neither pattern, by itself, proves redundancy.
Rank #4
The accompanying ILO explainer describes the results as potential exposure. Country infrastructure, technology adoption, available skills and employer choices determine whether a task is changed, supported or automated.
How to read the headline numbers
| Figure | How to interpret it |
|---|---|
| One in four workers worldwide | ILO 2025 estimate of workers in occupations with some degree of GenAI exposure; it is not a share of jobs predicted to vanish. |
| 3.3% of global employment | ILO 2025 estimate in the highest exposure category; the proportion varies by gender and national-income level. |
| Mean automation score 0.29 in 2025 versus 0.30 in 2023 | ILO assessment scores, not percentages of jobs lost. The reported standard deviation also fell from 0.30 in 2023 to 0.14 in 2025. |
| Eight-percentage-point increase in a vacancy measure | OECD’s 2024 analysis found the share of vacancies in highly AI-exposed occupations asking for at least one emotional, cognitive or digital skill rose by 8 percentage points. Establishment-level evidence in the same work indicated demand for those skills was beginning to fall, so the vacancy movement is not a guaranteed long-term trend. |
What determines whether a particular job changes?
Use these questions instead of a countdown of supposedly doomed occupations:
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minutePC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Best Value
- Digital content: How much of the role is processing text, images, code, numbers or other digital information that current systems can handle?
- Task concentration: Is exposure limited to a few duties, or spread consistently across most of the job?
- Human responsibility: Which decisions require professional judgement, legal accountability, safety oversight or a person willing to explain the result?
- Social and physical context: Does the work depend on trust, negotiation, care, collaboration, a physical site or changing real-world conditions?
- Work design: Can the employer redesign the process, verify outputs and absorb implementation costs?
- Local conditions: What do the country’s infrastructure, regulation, labour supply and customers permit?
How to assess your own role without relying on a job-title ranking
- List the recurring tasks. Write down a typical week’s activities, including preparation, communication, checking, decisions and follow-up—not just the title on your contract.
- Mark the digital and repeatable parts. Identify tasks based on standardised inputs and outputs, such as drafting routine text, searching records, reconciling data or producing a first-pass report.
- Separate assistance from substitution. Ask whether an AI output can be used only after a qualified person verifies it, or whether the whole task could be handed over with acceptable risk.
- Identify the irreplaceable responsibilities. Note work involving accountability, relationships, tacit knowledge, physical presence, unusual cases and decisions whose consequences must be owned by a person or regulated professional.
- Track changing requirements. Watch job postings, internal workflow changes and the tools your employer actually deploys. The ILO and OECD evidence supports preparing for a changing task mix, not assuming that any particular course guarantees job security.
What employers and policymakers need to manage
Exposure estimates are useful for targeting consultation and training, but they do not replace workplace evidence. Employers should examine how a system changes quality controls, staffing, responsibility and customer outcomes before counting a task as automated. Workers need a voice in those decisions, especially where monitoring or evaluation changes.
The ILO argues for managing the transition through social dialogue. Pawel Gmyrek, the ILO senior researcher who led the index, said in the organisation’s 20 May 2025 news item: “We went beyond theory to build a tool grounded in real-world jobs. By combining human insight, expert review, and generative AI models, we’ve created a replicable method that helps countries assess risk and respond with precision.” Janine Berg, an ILO senior economist, added: “It’s easy to get lost in the AI hype. What we need is clarity and context. This tool helps countries across the world assess potential exposure and prepare their labour markets for a fairer digital future.”
The answer to “Which jobs will AI kill?”
No defensible evidence currently identifies occupations that artificial intelligence will certainly eliminate, or gives a reliable date for their disappearance. Generative AI is most likely, in the near term, to change the tasks and skills inside many jobs—most visibly in clerical work and some highly digitised professional fields—while the scale of actual job loss depends on adoption, work design, human accountability and local labour-market conditions.
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.
