Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Probably not outright—but AI may change a substantial part of your work. The International Labour Organization (ILO) says one in four workers worldwide are in occupations with some exposure to generative AI (GenAI). Its 2025 assessment says transformation of jobs is more likely than wholesale redundancy because most occupations still include tasks requiring human input.
That statistic is not a prediction that you will be laid off. Exposure measures where AI capabilities overlap with work tasks; it does not show that an employer has adopted a system, that a role has disappeared, or when any change will occur. Your outlook depends on your actual tasks, your employer’s decisions, local labor demand and how work is reorganized.
As an Amazon Associate I earn from qualifying purchases.
What “AI exposure” means—and what it does not
A job title is a bundle of tasks. Some can be automated or accelerated by software, while others depend on judgment, accountability, relationships, physical presence, confidential context or decisions that an organization still assigns to a person. Exposure research estimates how much of that task bundle could be affected; it does not count confirmed layoffs.
The ILO’s refined 2025 index uses task-level assessments, worker input, expert review and AI-assisted predictions. It covers 29,753 tasks in the Polish occupational classification, drawing on 52,558 observations of perceived automation potential for 2,861 tasks. Occupations are grouped into four gradients of increasing exposure. The index is designed for cross-country analysis, not as an individual risk calculator.
#1 Best Overall
As ILO Senior Economist Janine Berg put it when the index was released on 20 May 2025, the aim is to add “clarity and context” so countries can prepare labor markets. The index measures potential capability overlap, not adoption or actual employment outcomes.
The headline ILO figures
| Measure | What the 2025 ILO index reports | How to interpret it |
|---|---|---|
| Global exposure | One in four workers are in occupations with some degree of GenAI exposure | A modeled share of employment in exposed occupations, not a share of jobs lost |
| Highest exposure category, worldwide | 3.3% of global employment; 4.7% of female employment and 2.4% of male employment | The most exposed gradient, still not a forecast of redundancy |
| Highest exposure category in high-income countries | 9.6% of female employment and 3.5% of male employment | Country income and occupational mix materially change exposure |
| Occupations with some exposure | 11% of employment in low-income countries versus 34% in high-income countries | Different digital infrastructure and job structures produce different modeled patterns |
| Mean automation score | 0.29 in 2025, compared with 0.30 in 2023 | An index score, not 29% or 30% of jobs eliminated |
| Score variability | Standard deviation fell from 0.30 in 2023 to 0.14 in 2025 | The refined method produced less variation in scores; it does not indicate falling layoffs |
The ILO reports that clerical occupations remain the most exposed. Highly digitized cognitive work in media, software, finance and technical roles has also become more exposed as GenAI systems improve. Two people with the same title can therefore face different levels of change if one handles routine document processing and the other handles clients, exceptions or regulated decisions.
Which jobs and tasks are most exposed?
Clerical and administrative work
Scheduling, standard correspondence, transcription, form completion, data entry, document classification and routine record updates are examples of tasks that fit current language and information-processing systems. The ILO has consistently found clerical work to have the greatest technological exposure. Automation of a portion of these duties may change staffing needs, but it does not establish that every clerical occupation will vanish.
Rank #2
Digitized professional and technical work
GenAI can assist with first drafts, code suggestions, research summaries, financial documents, media production and web content. The ILO’s 2025 index therefore shows increased exposure in some professional, technical, media, software and finance-related tasks. Exposure is not a ranking of which profession is “doomed”; it identifies where a larger share of work is performed on digital information that a model can process.
Work that remains difficult to automate completely
Tasks are less straightforward to automate when they require physical manipulation in changing environments, responsibility for safety, negotiation among people, tacit local knowledge, empathetic care, original goals, or a person who is legally and ethically accountable for the result. AI may still support these activities, but assistance and replacement are different outcomes.
What other evidence says about jobs and employment
The OECD’s 2024 analysis of vacancies across Austria, Belgium, Canada, Czechia, France, Germany, the Netherlands, Sweden, the United Kingdom and the United States found that about one-third of vacancies were in occupations highly exposed to AI. The range was 31% in Austria to 45% in the United Kingdom. Those are vacancy occupations with modeled exposure, not vacancies that disappeared.
The same OECD workplace report surveyed 5,334 workers and 2,053 firms in manufacturing and finance in Austria, Canada, France, Germany, Ireland, the United Kingdom and the United States. Around 80% of surveyed AI users said AI had improved their performance at work. That is a report of users’ perceptions in those sectors and countries, not a representative global employment result.
Free tools Windows power users keep installed
One-click scans. No signup required.
The OECD report also cited its 2023 Employment Outlook estimate that occupations classified as at highest risk of automation represented about 27% of total employment. “Highest risk” is a modeled possibility under broader automation, not an observed 27% job-loss rate. The report found little evidence at that time of a net negative effect on the number of jobs, while warning that displacement, unequal impacts and future adoption required monitoring. That 2024 assessment cannot guarantee future net employment.
Why the numbers cannot be placed on one scale
| Source and measure | Technology scope | What it describes |
|---|---|---|
| ILO 2025 index | Generative AI | Potential exposure of occupational tasks, using four gradients |
| OECD 2024 vacancy analysis | AI exposure in vacancy occupations | Share of vacancies in highly exposed occupations across 10 countries |
| OECD 2023 Employment Outlook estimate, cited in 2024 report | Broader automation | Occupations modeled at highest automation risk |
| OECD workplace survey | AI used by surveyed workers and firms | Worker-reported performance and workplace experiences |
Do not compare these percentages as if they measured the same thing. Check whether a figure concerns GenAI or broader automation, tasks or whole occupations, potential exposure or observed employment, the year and method, the country and sector, and whether it measures job quantity, job quality, skills or perceptions.
How to estimate your own risk without guessing
No official study cited here can predict when a named person’s job will disappear. You can, however, make a practical task-based assessment and then test it against evidence from your employer and local labor market.
- List your recurring tasks. For two weeks, record activities and approximate time: producing documents, searching, analyzing, communicating, operating equipment, resolving exceptions, supervising people and making decisions.
- Mark the information-processing tasks. Identify work based on repeatable digital inputs and outputs, such as standard reports, summaries, classifications, routine calculations or templated messages. These are candidates for AI assistance or automation, not proof of replacement.
- Separate routine steps from responsibility. Note where you must verify facts, interpret ambiguous context, protect confidential information, obtain consent, meet a regulation, or accept accountability. Those requirements can limit full automation even when a draft or recommendation is automated.
- Check the workflow, not just the tool. Ask whether your employer has approved systems, data access, security controls, training, procurement plans or targets for changing headcount. A capable public tool has no direct effect on your job until an organization chooses and can deploy it.
- Look for local signals. Review recent vacancies, internal role descriptions, collective agreements, professional guidance and changes in demand for your sector. National averages conceal differences between countries, industries and employers.
- Score the change in three ways. Classify each task as likely to be automated, likely to be augmented or unlikely to change soon. Then estimate whether the change saves time, raises expected output, shifts responsibility or removes a service altogether.
- Discuss the result with your manager or worker representative. Ask what systems are being considered, which tasks will change, how quality will be checked, what training is offered and how performance will be evaluated. Treat evasive answers as uncertainty, not as proof of imminent dismissal.
Skills that improve adaptability
The OECD recommends anticipating skill needs, training people to work with and develop AI, providing career guidance, targeting workers at direct risk and maintaining employment support and social protection. No single course, credential or purchase guarantees security. The useful goal is a combination of technical fluency and capabilities that remain valuable when tools change.
- AI and data literacy: understand what a model can and cannot do, check outputs, recognize hallucinations, protect sensitive data and document human review.
- Domain expertise: know the rules, customers, processes and failure modes that determine whether an output is usable.
- Analytical and cognitive skills: frame problems, test assumptions, compare evidence and make decisions under uncertainty.
- Communication and language skills: explain recommendations, write clear requirements and translate technical results for non-specialists.
- Social and emotional skills: negotiate, teach, collaborate, handle conflict and build trust—especially where decisions affect people.
- Management and implementation: redesign workflows, measure quality, coordinate change and establish accountability.
In exposed occupations, the OECD found demand for management, business, digital, emotional, social, cognitive and language skills. It also listed secretaries and administrative assistants, accountants and financial analysts, software developers, managers and human-resources professionals among highly exposed occupational groups in its vacancy analysis. Broader digitization and economic conditions may also be driving those skill changes, so AI should not be treated as the only cause.
A practical 90-day plan
- Days 1–30: complete the task inventory, learn your employer’s approved AI and data policies, and choose one low-risk workflow for supervised experimentation.
- Days 31–60: measure time saved and error rates, document where human review is essential, and ask for training tied to your actual role rather than generic tools.
- Days 61–90: turn the results into a portfolio of improved processes, updated responsibilities or cross-functional projects. At the same time, investigate external vacancies and career guidance so you have options if your employer reduces the role.
Why job quality matters even when a job remains
Employment effects are not limited to a binary question of employed or unemployed. The ILO’s 2023 analysis highlighted possible changes in work intensity and autonomy. A system that drafts more output may raise quotas, accelerate monitoring or reduce discretion rather than reduce headcount.
The OECD identifies risks involving displacement, unequal effects, bias, privacy, transparency and insufficient human oversight. Its workplace survey associated training and worker consultation with better worker outcomes, but that association does not prove that training alone prevents job losses. Consultation should cover data use, evaluation, appeal rights, safety, workload and who is accountable when an automated recommendation is wrong.
Technology transitions are shaped by human choices. Employers decide whether to use AI to remove tasks, expand services, reduce costs, improve quality or redesign jobs. Governments and worker representatives influence training, social protection and safeguards. As the ILO’s authors stated in 2023, “outcomes of the technological transition are not pre-determined”; people must guide the process.
What to do if your role is directly threatened
- Request a written explanation of which tasks and dates are changing, rather than relying on rumors about “AI replacement.”
- Ask whether redeployment, reduced hours, reskilling or internal vacancies are available before assuming dismissal is the only outcome.
- Document your domain knowledge, quality controls, customer relationships and measurable improvements; these show where human contribution remains essential.
- Use public employment services, professional bodies and targeted training to build a realistic alternative path. A career-change workbook can help organize options, but buying one has no demonstrated effect on employment outcomes.
- Review local legal, contractual and collective-bargaining protections, because consultation and notice requirements vary by jurisdiction.
The bottom line for your question
AI is more likely to change the task mix of many jobs than to erase whole occupations immediately. The strongest warning signs are a high share of repeatable digital tasks, an employer with a funded deployment plan and falling demand for the service those tasks support. The strongest protective factors are accountable judgment, human relationships, physical or contextual work, scarce domain knowledge and the ability to redesign processes with AI.
Use exposure statistics to identify questions, not to predict your fate. Map your tasks, verify what your employer is actually implementing, build skills across technical and human dimensions, and prepare a backup path using evidence from your occupation and location.
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.
