Goldman flags AI pressure in call centres and entry-level work

Goldman Sachs says AI is already weighing on labour markets in developed economies. Its 19 August 2026 research found US call-centre employment 39% below trend, with software publishing, management consulting and advertising services also below historical trends. Entry-level workers were particularly exposed.

Goldman estimated that about 300 million jobs globally are exposed to AI and modelled potential displacement of 6–7% of the US workforce over roughly 10 years. Those figures are exposure and scenario estimates, not a count of confirmed redundancies.

The findings align with Stanford Digital Economy Lab research revised on 12 August. It found no widespread economy-wide displacement through mid-2026, but employment among 22–25-year-olds in highly AI-exposed occupations was 19% below comparable peers. Stanford said the effect appeared mainly through reduced hiring and was concentrated where AI substituted for, rather than complemented, workers.

PwC’s June analysis of more than one billion job advertisements presents a mixed picture: AI-exposed firms recorded faster productivity, headcount and wage growth, while required skills changed more than twice as quickly.

Employer takeaway

For an Australian employer considering AI-related role reductions, establish whether the work—not merely the technology—has genuinely become unnecessary. Consult affected employees before deciding, assess redeployment into complementary or redesigned roles, and document the business rationale and alternatives before implementing any genuine redundancy process.