Calendar

Market calendar

Rate decisions, earnings and economic data in one place.

Next up
🔒 Members only

Never miss an important moment again

  • ▦
    The full calendarAll rate decisions, earnings and macro data, with sources.
  • ★
    Choose what you followOnly the topics you care about.
  • !
    Important newsWhat really matters, summarised briefly.
  • ⏰
    RemindersPut any moment in your own calendar in one click.
Join for free Free. Education, not advice.
AI

These 30 jobs are under pressure from AI

· 9 min read · By the BitFox newsroom

An employee watches as an automated system takes over tasks at a modern office workstation. AI image

Research and illustration supported by AI. Checked by the BitFox newsroom.

In short

  • One in four workers worldwide is in an occupation with some exposure to generative AI.
  • Administration, customer service, financial processing and digital office work face significant exposure.
  • The list covers 30 jobs with vulnerable tasks, not a prediction that all these jobs will disappear.
  • Within five years, working alongside AI is likely to become a standard part of many office roles.

By 2029, a company could have software handle much of its administration. It can read documents, check information and send messages without a person completing every step. The technology already exists. The real question is how quickly employers will redesign their workflows.

AI does not need to become smarter than every human to change the labour market. It only needs to perform certain tasks faster, more cheaply and well enough. Jobs built around processing information, writing routine content and answering standard questions face the greatest exposure. Which roles are most vulnerable?

AI can now perform tasks that once required specialist skills

Progress is moving quickly. Stanford's 2026 AI Index Report shows that AI has reached or exceeded human performance on several demanding knowledge tests. On the SWE bench Verified coding benchmark, performance rose from around 60 percent to almost 100 percent in one year.

There is still a major difference between a strong test score and working independently inside a real company. AI agents, systems that carry out multiple actions on a computer, achieved about 66 percent task success on the OSWorld benchmark in 2025. That is progress, but it also means they fail roughly one in three tasks.

AI is not yet a reliable replacement in every situation. But it is already useful for work with clear rules, repeatable steps and results that can be checked. That is where the labour market is beginning to change.

One in four workers has a job exposed to AI

In 2025, the International Labour Organization, or ILO, studied almost 30,000 individual work tasks. Its conclusion was that one in four workers worldwide is in an occupation with some exposure to generative AI.

The ILO expects most of these jobs to change rather than disappear completely. That distinction matters. Accountants, for example, may remain in demand while invoice entry, amount checks and routine reporting become largely automated.

European figures show that businesses are already adopting the technology. Eurostat reported that 20 percent of EU enterprises with at least ten employees used AI in 2025. In 2024, the figure was 13.5 percent. Among large companies, adoption reached 55 percent in 2025.

The transition has already started. But which tasks are most exposed?

These 30 jobs face the greatest pressure over the next three years

The following list is an assessment based on tasks that AI can already support or perform. It is not an official ranking, nor a prediction that every role will disappear before November 2029. The number of available positions and the content of each job are more likely to change.

Administration and office work

  1. Data entry clerk. AI reads documents and enters information into systems.
  2. Routine administrative assistant. Applications, forms and records can be processed automatically.
  3. Invoice processing clerk. Software compares invoices, amounts and payment details.
  4. Payroll assistant handling routine work. Repeated calculations and checks can be automated.
  5. Document management clerk. AI sorts, labels and searches large collections of files.
  6. Routine appointment scheduler. Digital assistants arrange meetings and send reminders.
  7. Basic reporting assistant. AI collects figures and prepares recurring reports.

Customer service and communication

  1. First line customer service representative. Chatbots increasingly answer standard questions.
  2. Email support agent. AI drafts replies and handles simple requests.
  3. Telephone support agent handling routine enquiries. Voice assistants can hold conversations and record information.
  4. Telemarketer handling standard sales calls. AI can manage large volumes of calls and follow up messages.
  5. Basic complaints processing clerk. AI reads complaints and assigns categories.
  6. Booking and reservation clerk. Systems can check availability and process bookings.

Finance, insurance and legal support

  1. Bank administration clerk. Many routine checks and transactions can be automated.
  2. Basic credit processing assistant. Software collects information and performs initial checks.
  3. Insurance claims administrator. AI can assess documents, photographs and claim forms.
  4. Standard insurance claims handler. Straightforward cases can be processed automatically.
  5. Junior legal administrative assistant. AI searches documents and prepares summaries.
  6. Standard contract reviewer. Software identifies unusual clauses and missing information.

Media, marketing and creative work

  1. Basic website content writer. AI creates product descriptions, routine articles and summaries.
  2. Proofreader handling standard content. AI finds language errors and rewrites sentences.
  3. Basic translator. AI increasingly translates general documents and messages.
  4. Entry level graphic designer. AI creates advertisements, banners and standard designs.
  5. Social media assistant producing routine content. AI creates posts, captions and publishing schedules.
  6. Basic market research assistant. AI gathers information and summarises competitors.

Technology, sales and human resources

  1. Junior programmer handling simple tasks. AI writes and checks more standard code.
  2. Manual software tester. AI runs repeatable tests and reports errors.
  3. First line IT helpdesk worker. AI resolves many common computer problems.
  4. Recruiter handling initial screening. AI compares CVs and schedules interviews.
  5. Sales support assistant. AI prepares quotes, finds customer information and processes routine requests.

Some of these roles may seem surprising. Programmers, designers and legal staff are not automatically protected because their work sounds specialised. Digital tasks with a clearly defined output are increasingly accessible to AI.

Young workers may lose opportunities first

A Stanford University study revised in August 2026 found an important early signal. Among US workers aged 22 to 25, employment in occupations highly exposed to AI was 19 percent below the level expected if it had followed the trend among less exposed occupations.

The researchers found no evidence of widespread job losses across the entire US economy. The gap was mainly linked to fewer new hires among younger workers. Experienced employees did not show the same pattern.

That distinction matters. Companies may hire fewer junior employees because AI can handle basic tasks. This makes it harder for young people to gain experience. And without a first job, moving into a specialist role becomes more difficult.

What does that mean? Junior employees often learn by completing simple assignments. If AI takes over those assignments, part of the route to becoming an experienced professional may disappear too.

This research concerns the United States. It does not prove that the same percentage applies to the Netherlands or Belgium. It does show where early pressure may become visible.

By 2031, AI may work like a digital colleague

The next step is likely to go beyond better chatbots. AI agents can carry out several actions in sequence.

An employee might ask an AI system to review a customer file. The system finds the information, compares documents, prepares a summary and drafts a response. The employee checks the result and makes the final decision.

That changes the job. One employee may be able to handle more cases. A company may therefore need fewer people to complete the same amount of work. At the same time, new roles may emerge in AI oversight, workflow management, security and handling exceptions.

The World Economic Forum expects 170 million new jobs to be created worldwide between 2025 and 2030, while 92 million existing jobs are displaced. That would mean a net increase of 78 million jobs. The estimate covers several forces, including technology, demographics and the green transition. It is not an AI only forecast.

The same report estimates that 39 percent of current worker skills will change by 2030. People who continue doing only the same routine tasks face greater exposure than those who learn to use AI and assess its output.

Some jobs are harder to replace than others

Work that combines personal contact, physical tasks and responsibility is harder to automate completely. Examples include nurses, plumbers, electricians, mechanics, cooks and childcare workers.

That does not mean AI will have no effect on these professions. A nurse may spend less time on paperwork. A mechanic may use AI to diagnose a fault more quickly. A cook may use software for stock management and scheduling.

The European Commission is also examining how AI changes work. Early European evidence points mainly to changes in tasks and occupations rather than a general wave of job losses. The outcome depends on costs, regulation, employers and whether customers continue to expect human contact.

Watch out: A job with high AI exposure does not automatically disappear. A role may even grow if AI makes employees more productive and demand for the service increases.

The question is not just which jobs disappear, but who benefits

If a company can complete the same work with fewer employees, staff costs may fall. That could increase profits or create room for lower prices. But the gains do not automatically reach workers.

For employees, the change can go in two directions. People who use AI well may take on more responsibility and complete more work. Those whose roles consist mainly of repetitive tasks may face fewer vacancies or changing job descriptions.

Governments will also have to respond. If work shifts from people to software and capital, the distribution of income from wages and business profits changes. That eventually affects tax revenue and social security systems.

Learning how to use AI is therefore not only for people who want to change careers. It is becoming useful in existing jobs too. Workers need to understand what AI does, where it makes mistakes and when human judgement is still required.

Questions to ask before worrying about your job

  • Which parts of my job can AI already perform?
  • How much of my work consists of fixed steps and standard information?
  • Can I use AI to do my work faster or better?
  • Which human skills, responsibilities or customer relationships make my role valuable?
  • Does my employer offer training to help staff work with AI?

Over the next three years, 30 entire professions will not automatically disappear. But the tasks they involve and the number of available positions may change. This article is for educational purposes only and is not financial or career advice.

Sources

Share: WhatsApp LinkedIn X
Note: This is education and not financial advice. Investing and crypto carry risk and you can lose your deposit.

The best explainers in your inbox every week

Also read