European business services and catering equities have felt sustained pressure over the past 16 months as investors weighed the potential for artificial intelligence to disrupt white-collar jobs. A sector report from Bernstein, published on May 15, 2026, challenges the depth of that market reaction and argues investors may be overstating the long-term employment risks posed by generative and agentic AI.
Market sentiment and the core argument
Bernstein notes a marked cooling of investor sentiment toward names judged to be heavily exposed to AI. The firm’s central contention is straightforward: AI should be seen primarily as a productivity-enhancing force that alters the nature of everyday tasks and speeds the introduction of new services, rather than as a mechanism that inevitably strips revenue from established corporate leaders.
The report highlights that, while automation will change how work is performed, historical precedent tends to show reallocation of labor across occupations instead of an outright collapse in employment levels. That observation informs Bernstein’s view that current stock-price weakness tied to long-term labor disruption risks may be an overreaction.
Heterogeneous effects across sub-sectors
Bernstein emphasizes the uneven exposure within the business services industry to the rollout of generative and agentic AI tools. Some parts of the sector face immediate pressures on pricing and volumes because basic tasks such as translation and customer service are relatively straightforward to automate. By contrast, segments that rely on manual, in-person work - notably blue-collar temporary staffing and physical testing firms - show structural resistance to automation because their revenue models depend on hands-on activities.
That divergence suggests the aggregate sector picture masks important differences in vulnerability and resilience across business lines.
Evidence and scenario framing
Bernstein references scenario work that projects significant automation of work hours. Under a cited 2030 midpoint scenario from McKinsey, roughly 30% of work hours in Europe and the United States could be automated. Even so, the expectation in the report is that net employment will tilt toward roles that are "augmentation-prone" rather than disappearing outright.
Supporting that framing, Bernstein points to corporate data from Adecco, where the company’s career transition business attributed only 1.4% of total corporate layoffs directly to AI integration. The report also underlines a strong historical association between labor productivity gains and employment growth - noting that in the U.S. the annual correlation between rising productivity and employment is about 80%, and that correlation reaches 100% when measured over a ten-year horizon.
Implications for investors and markets
The analysis suggests investors should differentiate within the business services universe and weigh productivity upside that may sustain revenues for incumbents against concentrated risks in data- and customer-heavy businesses. For some firms, the near-term operational effect of automation will be material; for others, particularly those grounded in manual services, structural immunity will limit disruption to revenues.