Meta Tests Robots to Handle Data Center Work

Meta is testing robots to maintain the data centers powering its artificial intelligence systems, including machines that can replace networking cables, restart servers, and inspect equipment.

According to a Friday WIRED report, Meta is testing robots from San Francisco-based Watney Robotics, Quebec-based Kinova, and Zurich-based ABB that could reduce labor costs—and potentially staffing—as its AI infrastructure expands.

One unnamed employee estimated that a successful cable-swapping robot could replace as much as 80% of the work performed in certain roles. The figure was the worker’s estimate, not a Meta projection, and the machine cannot yet match a person’s speed.

Robots enter the server room

Meta is testing robots to move server racks, inspect equipment, replace cables, and restart servers. However, the machines still need human supervision and struggle with obstacles, battery life, visual inspections, and dense cabling.

Nvidia and Alibaba are also developing systems to improve robot training. In June, Alibaba unveiled its Qwen-Robot Suite, with models for navigation, object handling, and physical simulation, while Nvidia researchers introduced AI agents that train robot fleets.

Limited real-world data, however, remains a barrier to bringing robots to the warehouse floor. ACE Robotics Chairman Wang Xiaogang said embodied AI—which enables machines to perceive and act in physical environments—could have a “ChatGPT moment” by the end of 2027.

Companies are also testing robots outside data centers. Bedrock Robotics has deployed autonomous construction equipment, and Mercedes-Benz has tested humanoid robots for repetitive factory work. Their developers cite labor shortages and worker safety, while critics warn that automation could eliminate jobs.

Jobs and the AI boom

Meta’s robotics push is fueling concern that the data centers built to power AI could eventually employ fewer people. Workers told WIRED that automation may reduce demand for experienced technicians and shift remaining tasks to lower-paid employees following AI-generated instructions.

A Meta spokesperson said the company continues to hire and train workers amid a skilled-labor shortage.

“America is in the middle of its biggest infrastructure boom since World War II, and there’s a major shortage of skilled workers to fill the roles; we need more workers, not fewer,” the spokesperson reportedly said.

That fear conflicts with the industry’s case for automation. Nvidia CEO Jensen Huang envisions data centers as “AI factories” operated by people, software agents, and robots. Huang and robotics developers argue that machines can fill labor shortages and handle repetitive or dangerous work.

On Wednesday, Microsoft co-founder Bill Gates proposed taxing robots and AI tokens, arguing that current tax rules can make machines cheaper than employees. Meta’s robots still need people to supervise them and perform difficult repairs, but their development could determine whether the company’s expanding data center footprint creates jobs or replaces them.

Meta did not immediately respond to a request for comment by Decrypt.