Analysis frame
Primary-source evidence
Interpret expectations as a measure of labor-market legitimacy and bargaining confidence, not as a prediction of aggregate employment outcomes.
- Young adults entering labor markets during rapid AI adoption
- Workers whose tasks are exposed but whose occupations may persist
- Employers asking staff to accept AI-led workflow change
- Governments designing training, income support, and labor-market reporting
- How expectations compare with actual employment, wage, and task changes
- Which events or information drive concern in each country
- How many respondents distinguish job transformation from net job loss
- Whether paid training and shared productivity gains would materially change public expectations
- Low trust may slow adoption even when a tool has measurable productivity value
- Political demand for income protection and employer disclosure may rise
- Firms may substitute contractors or hidden automation to avoid visible workforce commitments
- Public pessimism could become self-reinforcing if young people avoid exposed fields
The pattern spans very different economies
The survey covers high- and middle-income countries with different labor markets, education systems, welfare states, and exposure to frontier technology. The consistency of the direction is therefore more notable than any single national percentage.
At the same time, uncertainty is high in many places, especially where awareness is lower. The results capture a developing public judgment, not a settled global consensus.
Expectation is not outcome evidence
People can overestimate visible threats and underestimate new work that does not yet have a name. Net employment can also hide wage loss, deskilling, reduced entry-level opportunity, and geographic concentration.
The right comparison is therefore not survey fear versus one headline employment number. It is expectation versus observed changes in tasks, hours, wages, hiring, advancement, and who captures the productivity gain.
Legitimacy can be measured at the workplace
Employers can publish which tasks a system changes, whether positions disappear, how output and compensation move, and who receives paid training. Governments can require comparable disclosures and evaluate transition programs against actual placement and wage recovery.
That evidence will not eliminate conflict, but it can replace a vague promise of future jobs with a testable distribution bargain. Trust should rise only when workers can see the bargain in their own role.
Go to the source
Read the evidence behind this analysis. External links open in a new tab.
Pew Research Center — Global expectations for AI and employment


