BLS Projects 862,600 Fewer Jobs in 58 AI-Exposed Occupations

The U.S. Bureau of Labor Statistics has identified 206 occupations with very high AI exposure, including 58 projected to shrink by 2035. Together, those 58 occupations are expected to lose 862,600 roles, according to BLS analysis published on 27 August.

Examples include word processors and typists, data-entry keyers, telemarketers, order clerks, payroll clerks and file clerks. BLS also found that high exposure does not automatically mean contraction: some technology and programming occupations remain projected to grow.

The projections describe expected employment changes, not announced redundancies. A Panorama analysis reported by Livemint found that AI’s early labour-market effect has appeared more through slower hiring and fewer job openings than through mass layoffs. Stanford researchers likewise reported limited overall employment impact so far, with early hiring slowdowns among recent graduates in highly exposed occupations.

What employers should do

Use occupation-level projections to test where technology may remove tasks, but do not treat exposure alone as evidence that a role is redundant. Before deciding on redundancies, document the proposed operational change, consult affected employees as required, and assess reasonable redeployment or reskilling options. In Australia, a genuine redundancy requires that the job itself is no longer required and that consultation and redeployment obligations are addressed—not simply that AI is available.