Amazon says it has 1M+ robots in its warehouses, the most it's ever had and near the count of human workers, with ~75% of global deliveries assisted by robotics
The e-commerce giant now counts more than one million of the machines at its facilities — | Photographs and Videos by Desiree Rios for WSJ
Context & Ripple Effects
Amazon's warehouse automation has progressed from 45,000 robots across 20 fulfillment centers in 2017 to a portfolio spanning robotic arms, heavy-lift units and sorting machines, which had already surpassed 750,000 devices earlier in 2025. The new count shows that deployment is now approaching parity with Amazon's human warehouse workforce.
The milestone matters because Amazon has been moving robots into tasks alongside people, including AI-powered mobile robots that work alongside human workers. It also lands against evidence that warehouse employment can still grow even as automation spreads, because robots have struggled with some basic but consequential work.
First-order effects
- Amazon can route a larger share of fulfillment activity through robotic systems: it says roughly three-quarters of global deliveries now receive robotics assistance.
- Warehouse employees increasingly work in facilities where the robot fleet is nearly as large as the human workforce, making robot orchestration and human-machine handoffs more central to daily operations.
Second-order effects
- The scale raises the operational benchmark for competing retailers and logistics providers, increasing pressure to adopt automation where it can improve throughput, sorting or material movement.
- Demand shifts toward the hardware, software and facility-design capabilities needed to operate large mixed fleets, rather than stand-alone robots; Amazon's earlier expanding range of warehouse robot types illustrates that systems integration challenge.
Third-order effects
- If deployment continues, warehouse productivity will be defined less by headcount alone and more by how effectively operators combine workers with automated movement, picking and sorting systems.
- The labor effect remains non-linear: warehouse employment has risen despite automation in the broader sector, so this milestone points to job redesign and task reallocation before it establishes a simple one-for-one replacement pattern.
The trend: Large logistics operators are moving from isolated warehouse automation toward fleet-scale, human-and-robot operating models that make robotics a standard layer of fulfillment infrastructure.