«Computer Glitch» at New Jersey Data Center Triggers 5,000-Gallon Diesel Spill Amid AI Backlash

A major diesel spill at the Equinix NY2 data center in Secaucus, New Jersey, has added another controversy to the growing backlash surrounding AI infrastructure. Approximately 5,000 gallons of No. 2 diesel escaped from the facility on September 11, with some fuel reaching nearby Anderson Creek and triggering an emergency environmental response.

Secaucus Mayor Michael Gonnelli attributed the failure to a «computer glitch» affecting the facility’s automated fueling system. According to his account, the system repeatedly kept «calling for oil», continuing to pump diesel into an outdoor generator storage tank until it overflowed and fuel entered a storm drain connected to the nearby creek.

The escaped fuel reached Anderson Creek, a tidal tributary running through the environmentally sensitive New Jersey Meadowlands and connecting with the Hackensack River. State environmental officials said containment measures prevented the spill from reaching the river itself, although authorities have not established exactly how much of the approximately 5,000 gallons ultimately entered the creek.

Equinix contracted environmental remediation company Clean Harbors to respond to the accident. Crews deployed containment booms and absorbent materials while vacuum trucks removed contaminated water and diesel from the creek. Environmental advocates warned that fuel trapped in mudflats, sediment and marsh vegetation could pose risks to fish, crabs, other aquatic organisms and birds feeding throughout the wetlands.

The accident has drawn attention to infrastructure rarely associated publicly with artificial intelligence. Data centers require enormous and reliable electricity supplies, meaning facilities often maintain industrial backup generators and substantial fuel reserves for emergencies. Those systems are designed to keep servers operating through power failures, but the Secaucus accident demonstrates that they can also introduce conventional industrial and environmental risks.

Public resistance to data centers was increasing before the New Jersey spill. Communities have raised concerns about massive electricity requirements, water consumption, utility costs, noise and environmental impacts as companies rapidly expand computing capacity for AI. Recent national polling has found substantial opposition to new facilities, transforming what was once primarily a local development dispute into a broader political issue.

That backlash is becoming increasingly relevant ahead of the November midterm elections, when control of Congress is at stake. Trump’s longtime pollster Tony Fabrizio recently warned Republican candidates against appearing unconditionally supportive of data centers, advising them to recognize voter concerns surrounding electricity prices, water use and community impacts before emphasizing jobs and economic investment.

Trump himself has taken a dramatically different tone, aggressively defending rapid AI infrastructure expansion as necessary for American prosperity and technological leadership. He has portrayed resistance as harmful to competition with China, declaring: «There is a SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China. WHOEVER WINS AI, WINS!»

The data center dispute is unfolding alongside a wider argument about how quickly increasingly powerful artificial intelligence should advance. Some prominent AI industry figures have advocated stronger safeguards surrounding frontier systems, citing potential cybersecurity and other national-security dangers. Trump has resisted efforts he views as unnecessarily restricting American development, arguing that slowing domestic innovation could surrender technological advantages to China.

For communities hosting data centers, however, concerns increasingly extend beyond theoretical questions about future AI systems. These facilities require power infrastructure, cooling equipment, generators, fuel tanks and other industrial machinery. The Secaucus accident transformed those concerns into something tangible: thousands of gallons of diesel escaping after an apparent automated-system malfunction and contaminating a creek within an environmentally sensitive wetland ecosystem.

The national argument ultimately returns to Anderson Creek, where cleanup crews were left dealing with the physical consequences of the malfunction. While Trump argues that expanding data centers is essential to winning the global AI race, critics now have another incident to cite: a «computer glitch» at a New Jersey facility that allegedly kept demanding fuel until thousands of gallons escaped.