Americans displaced from jobs likely affected by artificial intelligence are experiencing setbacks resembling those associated with earlier technological shifts, according to Goldman Sachs. Since 2025, these workers have taken about a month longer to find new employment than other displaced workers, while their probability of remaining unemployed after a year was approximately 10 percentage points higher.
Age has emerged as an important dividing line. Workers 35 and older required about five additional weeks to return to full-time employment, compared with roughly three weeks for younger workers. Goldman's historical research, drawing on data extending back to 1979, found that technology-displaced workers experienced slower inflation-adjusted earnings growth and persistently greater risks of subsequent unemployment.
Separately, some employers are reversing staffing decisions after AI systems failed to meet expectations. Hiring managers surveyed by recruitment firm Robert Half reported restoring the same or similar roles after eliminating positions primarily because of AI or automation. Frequently cited reasons included unexpectedly high requirements for human oversight and an inability to reproduce institutional knowledge.
Klarna resumed recruiting human customer service representatives last year after relying too heavily on an AI chatbot to reduce costs. Ford said in June that 350 experienced engineers had been hired or promoted after the automaker's AI tools for quality control fell short of expectations.
Management experts advise returning employees to negotiate pay, training, advancement opportunities, and clarity about future AI-related staffing decisions. Georgetown University management professor Brooks Holtom also suggested seeking contractual protection, such as a renewable one-year agreement with compensation if the position ends early.
