Are any jobs AI-proof? What UK research shows
Updated · 5 sources
No study can promise that a job will stay the same. In a UK exposure ranking, several hands-on trades sit near the less-exposed end, while electricians sit below most groups but outside that lowest set. Exposure means a task could be affected by AI. It says little about demand, robotics or the choices an employer will make.
What the evidence says
Roofing and plastering
In the Department for Education's AI exposure ranking of UK occupation groups, roofers came 364th of 365 and plasterers 362nd of 365.[1]
Source: Department for Education (Unit for Future Skills) · 28 November 2023 · UK
Painting and bricklaying
On the same exposure ranking, painters and decorators came 353rd of 365 and bricklayers and masons 349th of 365.[1] Each rank describes an occupation group, not one person's job.
Source: Department for Education (Unit for Future Skills) · 28 November 2023 · UK
Electrical work
Electricians and electrical fitters came 246th of 365 for exposure: lower than most groups, but not in the least exposed set.[1]
Source: Department for Education (Unit for Future Skills) · 28 November 2023 · UK
An older model ranked them the other way round
Before generative AI, an ONS task model of jobs in England gave electricians an automation score of 0.51, above the middle score of 0.47 that we calculated from its data, and gave programmers 0.27.[2] In the newer AI exposure ranking those two groups swap places.
Source: Office for National Statistics · 25 March 2019 · England
Physical work in a London study
GLA Economics estimates that at least 46 per cent of London's workers, against 38 per cent across the UK, are in roles where generative AI could automate part of the work.[3] Its authors judge work done in person, including skilled trades and hands-on care, to be less directly exposed, while noting that no sector is likely to be untouched.
Source: GLA Economics (Greater London Authority) · 30 April 2026 · London
What it doesn’t tell you
- These are AI exposure ranks, not evidence that any occupation will keep its current number of jobs.
- The ranking does not cover robots performing physical tasks, and occupations contain different mixes of tasks.
- The government warns that exposure indices are indicative rather than precise, so small differences between ranks deserve little weight.
- Studies built for different technologies rank the same jobs in different orders, so one study is a weak basis for choosing a trade.
- The London estimate covers London, and its UK comparison uses an international exposure measure, not the Department for Education's.
DfE caveat: The analysis measures the exposure of jobs to AI, rather than distinguishing whether a job will be augmented (aided) or replaced (substituted) by AI.
Compare the work behind the ranking
Use a low exposure rank as a starting point for questions about the job. Ask a local employer about physical demands, travel, supervision and entry opportunities, then compare the training route with your circumstances. A rank cannot establish that a particular vacancy exists or that the work suits you.
Read the source: Department for Education (Unit for Future Skills)
If you’re weighing up a trade
Choose a trade by the work, its training route and its regulated responsibilities. Our workshop lets you try a small part of electrical learning before committing to a course.
We don’t teach roofing, plastering, painting and decorating or bricklaying yet
We currently teach electrical work. See the route into electrical work.
Common questions
- Are electricians among the least exposed?
- No. Their group is less exposed than most occupations in this study, but several practical trades are less exposed still.
- Did older studies also rank trades as less exposed?
- Not consistently. A pre-generative-AI ONS model scored many skilled trades above the middle for automation, because it looked at routine and repetitive tasks rather than language and analysis.
- Does a low exposure rank mean secure work?
- No. Technology, local conditions and policy can change work in ways the rank does not measure.
- Why teach electrical work first?
- It involves physical site work, legal responsibility and regulated sign-off. A low exposure rank alone would not justify a course.
Sources and checks
1.The impact of AI on UK jobs and training · Department for Education (Unit for Future Skills)
Published: 28 November 2023. Checked: 23 October 2026. Coverage: UK. Kind: exposure index. Licence: Open Government Licence v3.0.
- 364th of 365: roofers, roof tilers and slaters (UK, exposure index). Relative AI exposure across occupation groups; it does not predict replacement or an individual's prospects.
- 362nd of 365: plasterers (UK, exposure index). Relative AI exposure across occupation groups; it does not predict replacement or an individual's prospects.
- 353rd of 365: painters and decorators (UK, exposure index). Relative AI exposure across occupation groups; it does not predict replacement or an individual's prospects.
- 349th of 365: bricklayers and masons (UK, exposure index). Relative AI exposure across occupation groups; it does not predict replacement or an individual's prospects.
- 246th of 365: electricians and electrical fitters (UK, exposure index). Relative AI exposure across occupation groups; it does not predict replacement or an individual's prospects.
Supports: Occupation-group ranks and the source's caveat about interpreting exposure. Conflict: None declared in this source record. Data file
2.The probability of automation in England · Office for National Statistics
Published: 25 March 2019. Checked: 23 October 2026. Coverage: England. Kind: official statistic. Licence: Open Government Licence v3.0.
- 0.51: ONS automation score for electricians and electrical fitters, England, 2017 data (England, official statistic). A score from zero to one in a task model built before today's generative-AI tools; not a forecast.
- 0.47: median ONS automation score across the occupation groups it scored (England, official statistic). Our calculation of the middle value from the ONS occupation spreadsheet.
- 0.27: ONS automation score for programmers and software development professionals (England, official statistic). A score from zero to one in a task model built before today's generative-AI tools; not a forecast.
Supports: The pre-generative-AI automation scores for electricians and programmers in England. Conflict: None declared in this source record. Data file
3.London's workforce exposure to generative artificial intelligence · GLA Economics (Greater London Authority)
Published: 30 April 2026. Checked: 23 October 2026. Coverage: London. Kind: exposure index. Licence: © Greater London Authority; short citation only.
- at least 46 per cent: of London's workers in such roles (London, exposure index). London only, with more office and professional work than the UK as a whole.
- 38 per cent: of UK workers in roles where generative AI could automate a share of tasks (UK, exposure index). GLA Economics' UK average from the ILO's exposure measure; exposure is not job loss.
Supports: London and UK exposure shares, and the finding on physical, trade and care work. Conflict: None declared in this source record.
4.Assessment of AI capabilities and the impact on the UK labour market · Department for Science, Innovation and Technology and the AI Security Institute
Published: 28 January 2026. Checked: 23 October 2026. Coverage: UK. Kind: evidence review. Licence: Open Government Licence v3.0.
Supports: The limits of exposure indices. Conflict: None declared in this source record.
5.Open Government Licence v3.0 · The National Archives
Published: undated. Checked: 22 October 2026. Coverage: UK. Kind: licence. Licence: Open Government Licence v3.0.
Supports: Attribution terms for public sector research. Conflict: None declared in this source record.
Contains public sector information licensed under the Open Government Licence v3.0. No publisher endorses TradeTraining or this guide.
Changelog
- : Added practical questions for interpreting the evidence and structured data for the visible FAQs.
- : Added the ONS automation model, which orders electricians and programmers the other way round, and GLA Economics' London findings on physical work.
- : First publication comparing practical trades with the DfE ranking.