VarwoAI Changing UK STEM Jobs: Survey Evidence and Skills Planning
Read the UK STEM jobs discussion with the correct survey population, evidence limits and a practical way to assess changing work.
AI is changing tasks and review work in UK STEM roles. Current survey evidence does not establish that every technical occupation is disappearing or that an individual's job will be replaced.
Start with the survey population
SThree's 2026 STEM Workforce Report, checked October 31, 2026, describes a YouGov online survey of 4,126 STEM professionals in five markets, including 885 in the UK, fielded July 10–September 1, 2026. Its five-market figures are not automatically UK-specific estimates.
The report says 19% of UK AI users dropped a tool because verification took longer than the time saved. This denominator is AI users in the surveyed population, not 19% of all UK workers. It is self-reported experience, not a controlled productivity experiment or official vacancy count.
Energy Live News's October 22 report separately describes colleagues leaving after roles changed. Its “one in six” wording concerns respondents observing departures; it is not a measured fraction of all UK STEM jobs eliminated by AI.
Assess your own task changes
List one recurring task, its required expertise and who checks the result. For a proposed AI workflow, record drafting time, verification, correction and rework, then compare the complete task with the current approach. Keep examples where the tool fails.
| Role task | Capability worth checking |
|---|---|
| Analysis or coding | Validation and reproducibility |
| Engineering documentation | Traceability to requirements |
| Scientific writing | Evidence quality and source checking |
| Junior learning tasks | Practice with expert feedback |
This is a planning framework, not a forecast of earnings or employment. Varwo has not surveyed employers or run a workplace trial. For the research side of the discussion, read AI's theoretical-computer-science impact.