Yale research finds AI is shifting jobs more than eliminating them

Yale SOM research published on 25 September found modest employment effects from AI, rather than broad-based displacement. AI-exposed occupations, including business, financial and engineering roles, experienced roughly a 2% decline in their employment share. Firms adopting AI nevertheless grew their workforces through productivity gains and reallocating tasks, the researchers said.

The research attributed about 14% of occupational shifts to AI, indicating that changing job content and moving work between roles remain more common than outright elimination.

A Stanford Digital Economy Lab analysis covering 41 countries reached a similar conclusion: it found no evidence of economy-wide displacement from generative AI adoption, alongside senior employment growth of 6.7% at AI-adopting firms and modest productivity gains.

For employers, the findings support planning around tasks and capabilities rather than assuming every AI-affected role is redundant. Before proposing redundancies, identify which duties are genuinely no longer required, consult affected employees before deciding, and assess suitable redeployment or reskilling options. Any Australian process must still meet the genuine-redundancy requirements.