The question everyone is quietly asking, is my job safe? Now has a structured answer, at least by occupation type.
Using the UK Office for National Statistics Standard Occupational Classification (SOC) 2020, I’ve rated all 26 sub-major occupation groups on a 1–10 scale for how likely AI is to substantially take over the core work of that role. A score of 1 means very safe; 10 means very high risk of automation.
The SOC 2020 is the definitive UK framework for categorising jobs, used by the ONS for labour market statistics, salary surveys, and employment research. Grouping at the sub-major level gives a meaningful picture without getting lost in the detail of hundreds of individual job titles.
A few themes stand out:
- Administrative and secretarial roles sit at the top of the risk table. Structured, repetitive, document-heavy work is exactly what current AI tools do well.
- Skilled trades and caring roles are among the safest. Physical adaptability, unpredictable environments, and genuine human connection are challenging to automate.
- Professional occupations are mixed, AI is a powerful assistant in many of these roles, but deep expertise, judgement under uncertainty, and client trust keep humans in the loop for now.
The ratings are illustrative assessments based on the nature of the work, not econometric forecasts. Use them as a starting point for thinking about where your role sits on the spectrum, and what skills might be worth developing.
Caveat: AI is changing things so fast month on month now that you should reassess your positioning based on the latest information regularly.
Click any bar for a quick summary, or use the table below for the full keyboard-accessible dataset.
This table mirrors the current filter and sort selection, including the rationale for each score.
| SOC | Occupation group | Risk | Rationale |
|---|