Amazon Employees Report AI Tools Increasing Workload Rather Than Reducing It
Key Takeaways
- ▸Amazon employees report that internal AI tools are error-prone and require significant manual correction, negating productivity gains and increasing overall workload
- ▸ActivTrak's analysis of 163,638 employees across three years shows AI adoption universally increased work activity in all measured categories, contradicting promises of reduced workloads
- ▸The pattern reflects a broader economic reality where AI serves as a productivity multiplier for corporate profit rather than a tool for worker relief, mirroring unfulfilled promises made by previous technologies like social media and mobile phones
Summary
Amazon corporate employees have told The Guardian that the company's push for staff to use internal AI tools is counterproductive, with workers reporting that the "half-baked" systems often make mistakes requiring manual correction and verification, ultimately adding to their workload rather than reducing it. A broader industry survey by workforce analytics company ActivTrak analyzed data from over 163,000 employees across 1,111 organizations and found that AI adoption is actually increasing workloads across the economy, contradicting the Silicon Valley narrative of AI-enabled efficiency gains. According to ActivTrak's research, workers saw increases in nearly every measured work category after AI adoption, including a 104% increase in emails, 145% increase in chat messaging, and 94% increase in time spent on business management tools. Rather than replacing existing work, AI is being used as an additional productivity layer that fills any freed-up time with more tasks, ultimately benefiting companies seeking greater output while failing to ease the burden on employees.
Editorial Opinion
The disconnect between AI's promised benefits and workplace reality reveals a fundamental misalignment between corporate incentives and worker welfare. While technology companies market AI as a path to liberation from burnout, the evidence suggests it's being weaponized as another lever for extracting more output from already overextended workforces. This pattern—where efficiency gains are immediately recaptured by management rather than returned to employees—underscores why meaningful AI policy must prioritize worker protections alongside innovation.
