Meta is testing robots from vendors like Watney Robotics, Kinova, and ABB to plug in cables and power cycle servers inside its data centers, according to current and former workers, as Big Tech confronts a shortage of skilled personnel and surging compute demand amid the AI boom.
Robots Move Onto the Data Center Floor
Big Tech’s race to scale artificial intelligence infrastructure has pushed technology out of the server rack and onto the physical floor. Meta is actively experimenting with robots that can plug in cables, reset servers, and handle other tasks inside its data centers. The ongoing effort, which involves hardware from vendors including Watney Robotics, Kinova, and ABB, may eventually allow the company to operate its rapidly expanding infrastructure footprint with fewer human workers, keeping labor costs in check as spending on AI infrastructure accelerates.
In one experiment detailed by current and former workers, the company is evaluating whether a Kinova Gen3 robotic arm could handle power cycling or cutting off electricity to servers. Other evaluations involve specialized robots built to swap networking cables. At some facilities, the company has quietly adopted a simpler device that resembles a finger—or a pointy stick—designed to press the power button on a Mac Mini or other hardware to restart it when remotely prompted by a human worker.
One data center worker estimated that a successful deployment could replace up to 80 percent of some employee workloads. That prospect has triggered anxiety among technicians on the ground. We thought those of us performing the physical tasks were safe for a while, but not anymore,
one worker told the reporting outlet, adding, It’s coming for us all, unfortunately.
Meta declined to comment on the testing. Company spokesperson Francis Brennan pointed to heavy investments in training and hiring workers to build and operate the facilities. 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,
Brennan said in a statement.
At a conference last year, Eric Xu, senior manager for robotics at Meta, outlined the long-term vision. He argued that robots could speed up incident response, monitor environmental conditions, and handle preventative maintenance. We believe more collaboration and research will be needed, but we have to start, otherwise we don’t have a chance to do this,
Xu said.
A Workforce Crisis in Mission-Critical Operations
The push toward automation arrives as data center operators face a talent shortage. The construction industry alone faced a shortage of about 439,000 workers as of late 2025, mostly in the skilled trades that data center projects rely on.
Gigawatt-Scale Expansion and the New Geography of Power
Keeping pace with global demand for compute power could require $6.7 trillion in data center investment by 2030, according to analysis from Aligned Data Centers CEO and Board Member Andrew Schaap. That capital intensity mirrors major historical industrial buildouts.
Geopolitical Headwinds and Domestic Activism
As the digital infrastructure boom reshapes local economies, it has encountered a wave of domestic opposition and regulatory friction. According to recent polling, roughly three-quarters of Americans oppose the construction of AI data centers, driven by concerns over electricity, water consumption, noise, and local economic benefits.
A report by Sam Lyman, a former senior adviser and chief speechwriter to Treasury Secretary Scott Bessent and AI researcher who now leads research at the Bitcoin Policy Institute, attributes 10 data-center moratoria, one permanent ban, and four rejected or abandoned AI projects to activist campaigns tied to networks with documented links to the Chinese Communist Party. The report notes that organizations affiliated with Shanghai-based expatriate Neville Roy Singham have promoted narratives that align with Beijing’s interest in slowing America’s technological progress while supporting domestic policies that constrain U.S. infrastructure development.
Those advocacy efforts have found legislative echoes, such as the Sanders-Ocasio-Cortez Artificial Intelligence Data Center Moratorium Act, which seeks to halt qualifying projects until federal safety, environmental, and community mandates are enacted.
Adapting Thermal Demands and Resource Realities
Inside the facilities, engineering teams are racing to solve the physical constraints generated by high-performance computing.
Unlike finite resource extraction, modern digital infrastructure must continually adapt. Whether through advanced thermal cooling labs or experimental maintenance robots, the survival of the sector depends on balancing unprecedented computational appetite against finite physical resources and a strained labor market.
