Last updated: 23 November 2026
Quick answer: For most UK workers, AI is more likely to change what you do at work than to remove your job outright, at least in the near term. Roles made up mainly of routine, screen-based tasks with clear right answers are most exposed to automation. Hands-on, relationship-based and accountable work is least exposed. Assess your own job task by task.
Ask "will AI take my job?" and you will find answers ranging from "almost certainly" to "never". Neither helps you plan. This guide uses UK government research to show which roles are most and least exposed, explains the difference between automation and augmentation, and gives you a simple test to apply to your own work.
Key facts at a glance
Most exposed occupations: professional roles in finance, law and business management, such as management consultants, accountants and solicitors (Department for Education, November 2023).
Least exposed occupations: manual work in unpredictable settings, such as roofers, plasterers and steel erectors.
Most exposed sector: finance and insurance.
Regions: exposure is highest in London and the South East and lowest in the North East, but regional differences are much smaller than differences between jobs.
Global context: the IMF estimated that about 60% of jobs in advanced economies may be affected by AI, with roughly half of those likely to benefit.
UK adoption: around 1 in 7 UK employers (14%) said their site used AI in the Employer Skills Survey 2024.
Will AI take my job? The honest answer
Probably not all of it, but it may take some of your tasks. That distinction matters. Most research measures exposure: how much of a job involves tasks AI can help with. Exposure is not the same as replacement. A job can be highly exposed and still grow, if AI makes the people doing it more productive.
The Department for Education's 2023 analysis says its results "should be interpreted with caution", because exposure depends on uncertain assumptions and AI is developing quickly. That is the right attitude. Anyone who tells you exactly what will happen to your role by 2030 is guessing.
The IMF's January 2024 analysis makes the same point at a global level. In advanced economies such as the UK, it estimated about 60% of jobs may be affected. For around half of those, AI could raise productivity. For the other half, AI may carry out key tasks people now do, which could mean lower demand, lower pay or fewer hires in those roles.
Which jobs are most exposed to AI in the UK?
The Department for Education mapped the abilities each UK occupation needs against what common AI applications can do. Its findings challenge the idea that AI mainly threatens low-paid work.
Most exposed to AI overall
Management consultants and business analysts
Financial managers and directors
Chartered and certified accountants
Psychologists
Actuaries, economists and statisticians
Solicitors and other legal professionals
Most exposed to large language models (tools like ChatGPT)
Telephone salespersons
Solicitors
Further and higher education teaching professionals
Credit controllers
Human resources administration
Public relations professionals
Least exposed
Roofers, roof tilers and slaters
Elementary construction occupations
Plasterers and steel erectors
Fork-lift truck drivers
Gardeners and street cleaners
By sector, finance and insurance was most exposed, followed by information and communication, professional services, property, public administration and education. Accommodation and food, motor trades, agriculture, transport and construction were least exposed.
One finding surprises people. Employees with a degree-level qualification were more likely to be in highly exposed jobs than those with A-level equivalents. Qualifications in building and construction, manufacturing technologies and transport operations at level 3 or below were linked to the least exposed jobs.
What is the difference between automation and augmentation?
Automation means AI does a task instead of a person. Augmentation means AI helps a person do the task faster or better. The same tool can do both, depending on how an employer uses it.
The DfE report maps research from the International Labour Organization onto UK jobs. It lists occupations with high automation potential, where AI could substitute for a large share of tasks. They include:
Call and contact centre occupations and other customer service roles
Bookkeepers, payroll managers and wages clerks
Bank and post office clerks, and pensions and insurance clerks
Human resources and financial administrative occupations
Typists, travel agents and telephone salespersons
Authors, writers and translators
Professional roles such as accountants and solicitors are highly exposed but are more often described as candidates for augmentation. AI drafts, summarises and checks, while a qualified person still takes responsibility for the result.
How exposed is your job? A task-by-task test
Job titles hide too much. Two "administrators" can do very different work. Use this test on your actual week.
List the 8 to 10 tasks that fill most of your working week.
Estimate the percentage of time each one takes.
For each task, ask three questions:
Are the inputs and outputs mostly text, numbers or images on a screen?
Is it repetitive, with a clear right answer?
Would a mistake be cheap and easy to spot?
Three "yes" answers means high exposure. One or none means low exposure.
Add up the time spent on high-exposure tasks.
If high-exposure tasks take more than half your week, plan now. If they take less than a quarter, your priority is learning to use AI tools for those tasks so you get the time back.
Worked example: a payroll administrator
Example (composite, not a real person): Sarah, 42, is a payroll administrator for a care provider in Sheffield. Her audit looks like this.
Entering timesheet data: 30% of her week. High exposure.
Answering routine pay queries by email: 20%. High exposure.
Checking exceptions and errors before each pay run: 20%. Medium.
Handling sensitive cases, such as sickness, maternity pay and overpayments: 20%. Low.
Liaising with managers and HMRC queries: 10%. Low.
Half her week is highly exposed. Sarah's response is not to panic but to move towards the low-exposure work. She asks to take on more complex cases, studies for a payroll qualification and learns how her payroll software's automation features work, so she becomes the person who configures and checks them.
What is happening in UK workplaces now?
Adoption is real but uneven. In the Employer Skills Survey 2024, 14% of UK employers said their site used AI, and around two thirds of those had started in the previous 12 months. Use was much higher in information and communications (43%) than in construction (7%), and higher among larger sites.
That suggests change is arriving faster in some sectors and employers than others. A large bank and a small building firm are on very different timelines. It also means your employer's choices matter more than national averages.
What could the next five years look like?
No one can say for certain. If current trends continue, these are the signals worth watching in your own role.
Pilots: your employer testing AI tools on tasks you do.
Job adverts: roles like yours starting to ask for experience with specific AI tools.
Entry hiring: fewer junior roles in your team, as covered in our guide to how AI is changing entry-level jobs.
Workload shifts: fewer people doing the same volume of routine work, with more time on exceptions.
Recruitment: AI being used to screen applications in your field. Our guide on how UK employers use AI to screen candidates explains your rights.
Myths and facts about AI and jobs
Myth: AI only threatens low-paid jobs. Fact: UK government analysis found professional roles in finance, law and consultancy among the most exposed, while many manual jobs were among the least exposed.
Myth: high exposure means your job will disappear. Fact: exposure measures how much AI could help with a job's tasks. Many exposed roles are expected to be augmented rather than replaced.
Myth: a degree protects you. Fact: degree holders were more likely to be in highly exposed jobs. What protects you is the mix of tasks you do and how well you adapt.
Myth: if you work in a trade, you can ignore AI. Fact: quoting, scheduling, invoicing and customer messages are all screen-based tasks AI can speed up, even in low-exposure jobs.
Myth: everyone is already using AI at work. Fact: only around 1 in 7 UK employers reported using AI in 2024, so many workplaces are at an early stage.
What should you do about it?
Your action plan depends on your audit result.
If most of your tasks are highly exposed, start building skills for an adjacent role now, while you are employed. Look for roles that supervise, check or configure the systems doing your current tasks.
If your exposure is mixed, learn to use AI tools for the routine parts so you can spend more time on judgement and relationships. Our guide to AI skills for non-technical jobs is a good starting point.
If your exposure is low, do not ignore AI entirely. Basic confidence with the tools is becoming a general workplace skill.
If you are job hunting, learn how to use AI well without sounding generic. See our guide to using AI in your job search.
What to do next
Run the task-by-task test on your current job this week.
Pick the highest-exposure task and try doing it with an approved AI tool, if your employer allows it.
Identify one low-exposure task you could take on more of.
Ask your manager how the organisation plans to use AI in your area.
Browse short courses on Qamla to start building one adjacent skill.
Frequently asked questions
Will AI take my job in the next five years?
For most UK workers, the evidence points to AI changing parts of jobs rather than removing whole jobs, at least in the near term. Roles made up mostly of routine screen-based tasks, such as some admin, contact centre and bookkeeping work, face more risk. Nobody can predict your role precisely, so assess its tasks.
What jobs are safe from AI?
No job is completely unaffected, but UK government research found the least exposed roles involve hands-on work in unpredictable settings, such as roofers, plasterers, steel erectors and elementary construction jobs. Roles built on physical presence, care, trust and accountable judgement are also harder to automate than desk-based tasks with clear right answers.
What jobs will AI replace?
Research mapped to UK occupations lists roles with high automation potential, including call and contact centre work, bookkeeping and payroll clerks, typists, telephone salespersons, travel agents and some HR and finance administration. High potential does not mean those jobs will disappear, but routine tasks within them are most likely to be automated first.
Are graduates and professionals more at risk from AI?
They are more exposed, but not necessarily more at risk. The Department for Education found degree holders are more likely to work in jobs highly exposed to AI, such as consultancy, accountancy and law. Much of that exposure is likely to be augmentation, where AI speeds up tasks, rather than replacement.
How can I protect my job from AI?
Learn to use AI tools well in your current role, move towards tasks that need judgement, relationships and accountability, and keep evidence of the value you add. Watch for signals such as your employer piloting AI on your tasks. If most of your work is highly exposed, start building skills for an adjacent role now.
Is AI already causing job losses in the UK?
AI adoption is still uneven. In the Employer Skills Survey 2024, around 1 in 7 UK employers said their site used AI, with much higher use in information and communications than construction. Separating AI from wider economic factors in job figures is difficult, so treat strong claims either way with caution.
Sources and further reading
The impact of AI on UK jobs and training, Department for Education, 28 November 2023.
AI will transform the global economy. Let's make sure it benefits humanity, International Monetary Fund, 14 January 2024.
Employer Skills Survey 2024: full UK research report, Department for Education and IFF Research, November 2025 (updated June 2026).
This article summarises published research and is not a prediction about any individual job. AI is developing quickly, so check the latest sources before making major career decisions.
