Labourers in mining, construction, manufacturing and transport and AI: what the data say

What it means

The ILO index measures what share of an occupation’s tasks a language model could perform or speed up. It is technical exposure, not a layoff forecast: according to the index’s authors, a change in the content of work is more likely than the job disappearing.

The score is an unweighted average of the occupations in the group, at the global level rather than for Israel.

What is happening in Israel

Broad group in the Labour Force Survey: Elementary occupations

Relative risk: the group’s share of the unemployed divided by its share of the employed. Above 1 means over-representation among the unemployed. Source: CBS, Labour Force Survey, table 5.3. 2023–2024 include the war; the figure is descriptive.

What the research found

In 13 European countries there is still no significant link between firms’ AI adoption and employment in exposed occupations. The sign is negative, but the data cannot tell it apart from zero.

In Israel, clerical workers, the most exposed group, were over-represented among the unemployed before generative AI tools. On average in 2023–2024 the over-representation did not grow (relative risk 1.34, against 1.35 in 2017–2018); in 2024 alone it is 1.42. The clearer change is a rising share of professionals.

The full European finding → · Results chapter →

Occupations with similar exposure

• Market-oriented skilled forestry, fishery and hunting workers Rank 34
• Food preparation assistants Rank 35
• Subsistence farmers, fishers, hunters and gatherers Rank 37
• Building and related trades workers, excluding electricians Rank 38

All 40 occupation groups →