Before AI Replaces Jobs: What Five Enterprise Cases Teach About Role Redesign, Reskilling and Hiring

Artificial intelligence changes work inside a job before its full employment effects can be understood from headcount alone. This Evidence Brief compares Salesforce, DBS, PwC US, LONGi and JD.com to examine how task change, workflow redesign, decision rights, capabilities, learning experiences and career pathways reshape the architecture of a role.

The cases reveal different transition mechanisms: routine support moving to agents at Salesforce; banking roles linked to reskilling and internal mobility at DBS; audit work and graduate apprenticeship changing at PwC US; visual inspection expertise moving into AI training at LONGi; and customer-service knowledge becoming AI-operations expertise at JD.com. These are five selected enterprise cases, not a representative sample of the economy, and their operating metrics cannot be combined into a single estimate of jobs replaced by AI.

The report distinguishes task substitution from job elimination, training participation from successful transition, new role categories from net job creation, and reduced labour input from documented layoffs. It considers redeployment, attrition, non-backfill, reduced hiring and changed entry pathways as separate workforce effects. For employers, HR leaders, educators and policymakers, it offers a decision framework for redesigning roles, protecting learning opportunities and measuring the quality of worker transitions rather than relying on adoption statistics alone.