Will AI Take My Job? How to Check Your Own Risk

Will AI take my job? For most people the answer in 2026 is no, but it will change what your job contains. The International Labour Organization (ILO) finds that one in four workers worldwide hold a job with some exposure to generative AI, and it calls transformation the most likely outcome. So far the damage shows up at the entry door. Stanford researchers found that employment of 22 to 25 year olds in the most AI-exposed occupations stood about 19% below trend by June 2026. Your risk depends on your tasks, and you can check them yourself in an afternoon.

I train teams to work with AI and speak about careers and AI at work at company events. I build my own companies with the same tools I teach. The figures below come from the ILO, the World Economic Forum (WEF) and the Stanford Digital Economy Lab, checked in October 2026.

What the 2026 research says

Read these numbers with care. The ILO stresses that its figures measure potential exposure, not actual job losses. The Stanford team calls its results descriptive patterns, not proof that AI caused them. Even so, all three sources point the same way: AI is reshaping tasks inside jobs, and it hits the people trying to get in first.

The Stanford finding also shows the mechanism. The gap for young workers grew from 15% in the July 2025 data to 19% by June 2026. It came mostly from companies hiring fewer young people, and less from firing existing staff. The declines cluster in occupations where AI automates tasks. Where people use AI as a helper, employment stayed flat or grew, above all for experienced workers.

Will AI take my job? A five-step self-check

A job title hides too much. Two accountants can do very different weeks, so check your own tasks.

1. List your tasks

Write down what you did last week, task by task, with rough hours. Twenty lines is a good start. If you get stuck, look up your occupation in O*NET (opens in a new tab), the US occupational database with over 900 occupation profiles. Each profile lists typical job duties, and O*NET links to ESCO, the EU's classification, if you work in Europe.

2. Mark codified and tacit tasks

The Stanford team found a useful split. Codified knowledge is formal and documented: you could learn it from a textbook or a written procedure. Tacit knowledge comes from practice, mentorship and repeated real situations. Employment of young workers fell in jobs built on codified knowledge. It rose for experienced workers in jobs that rely on tacit knowledge.

Mark each task C or T. Filling a standard form is C. Calming an angry client who threatens to leave is T.

3. Give the tasks to an AI tool

Take five of your most frequent tasks and try them in ChatGPT, Claude or a similar tool this week, on real material. Before you paste anything, replace names, addresses, phone numbers and emails with placeholders. Germany's Federal Employment Agency recommends this placeholder method in its 2026 application guide. Use it for any work document.

Sort each result:

4. Count your hours

Add up the hours from step 1 for each group. A week that sits mostly in "done, usable as is" is the warning sign. A week that sits mostly in "needs your checking" means AI makes you faster, and the Stanford data suggest that kind of job holds up better.

5. Look at who gets hired into your role

Stanford found the adjustment happening through hiring. So ask: does your team still hire juniors for the work you started with? If the answer is no, the ladder you climbed is shortening, even if your own seat looks safe today.

Jobs AI will replace first, and the myth of safe jobs

The WEF asked over 1,000 employers, covering more than 14 million workers in 55 economies. They expect the largest declines in absolute numbers among clerical and secretarial workers, including cashiers, ticket clerks, administrative assistants and executive secretaries. The fastest-declining roles also include postal service clerks, bank tellers and data entry clerks. The ILO data agree: clerical occupations have the highest exposure.

The fastest-growing jobs in the same report are technology and green roles: big data specialists, fintech engineers, AI and machine learning specialists, software developers, renewable energy engineers and environmental engineers.

If you hoped manual work is safe from AI, the WEF data complicate that. Employers expect manual dexterity, endurance and precision to show a net decline in demand, and 58% of them expect robotics and automation to transform their business by 2030. Use the self-check above, and treat any "safe jobs" list as a starting point for research.

If your score is high

Move toward the tasks AI helps with. Look for the parts of your job that need judgement, relationships or experience in messy situations. Ask for more of them. A payroll clerk who takes over the conversations with employees about errors keeps the T tasks while the C tasks shrink.

Learn the tools inside your own work. The WEF puts AI and big data at the top of the fastest-growing skills. Two-thirds of the employers in the survey plan to hire talent with specific AI skills, and 40% expect to reduce staff where AI can automate tasks. If the world's workforce were 100 people, the report counts 59 who will need training by 2030. Learning the tool on your own documents beats a generic course, because you see at once what it changes in your week.

If you are early in your career, collect tacit experience fast. Internships, client-facing projects and work with a mentor build the kind of knowledge that held its value in the Stanford data.

Plan a move if your role is shrinking. A career change is easier while you still have a job. My guide on how to change careers walks through the steps I use in coaching.

What I see in the training room

In my workshops, each team works on its own documents and its own clients, and people leave with prompts and workflows they use the next day. That format follows the same logic as the self-check: start from your tasks, test the tool on them, keep what works. In Lithuania I run this training through my company Retos galimybės (opens in a new tab). In English, I run AI training for teams on site in Vilnius or remote.

Bring this to your team

I give talks on careers and AI at work that walk a room through the research above and a live version of the self-check. If you run an event or a team offsite, see my speaking topics or book a call and tell me who sits in the room. If you are an individual weighing a move, my career coaching starts with a 30-minute call.