GenAI could transform one in four jobs, and women carry more of the risk

Roughly one in four jobs worldwide falls into the zone exposed to generative AI.

The dominant effect is jobs being transformed at the task level, not eliminated outright.

Clerical and administrative work is the most exposed category.

In high income countries, about 9.6% of women's jobs are highly exposed, against about 3.5% of men's, because women are over represented in the most exposed roles.

Exposure is not destiny. Whether a transformed job becomes a better job or a lost one depends on whether the person builds the new capability the role now needs.

The gender gap is a fairness issue and a targeting signal. The most exposed roles are where capability building pays off first and matters most.

Map which roles in your organization are most exposed, then put capability support there first rather than spreading it thin.

Frame the change as roles being rebuilt, and give people a concrete path into the rebuilt version of their own job.

Start with the most exposed function, often administrative and clerical work, and redesign one workflow with the people who do it.

Track who is getting capability support so it does not quietly skip the roles that need it most.

Exposure indices estimate potential, not outcomes. A highly exposed job is not a job that will disappear.

The gender figures are strongest in high income countries and reflect job composition. Read them as a pattern to act on, not a verdict on any individual.