Amazon’s aggressive effort to infuse AI into white-collar work is drawing internal backlash from employees who say mandated AI use is slowing them down and increasing errors. Staff across engineering and research roles describe half-baked internal tools that hallucinate code, introduce bugs, and require extensive rework—undercutting leadership’s claims of faster delivery. The company has laid off roughly 30,000 corporate workers in recent months while urging remaining staff to adopt AI broadly, with managers tracking weekly usage on internal dashboards and, employees say, weighing AI enthusiasm in promotions. Amazon disputes that AI is hurting productivity, saying its tools have saved hundreds of millions of dollars and “thousands of years” of developer time, and that training and adoption are improving outcomes. The push mirrors a broader industry shift as companies including Block, Pinterest and Autodesk tie restructuring and investment to AI. Academics warn rushed deployments can expand surveillance and stifle skill development, especially when workers are tasked with “training” systems that could replace parts of their jobs. Recent service outages reportedly linked to internal AI use have heightened concerns. As one of America’s largest employers, Amazon’s approach could set norms for how AI reshapes management, performance measurement and job design across industries.
Related article:
