Amazon Web Services CEO Matt Garman has published a blog post of more than 3,000 words arguing that communities that block AI data center projects risk "irreparable harm" to the US economy and national security.

The post, published today on Amazon's corporate news site, represents the company's most direct attempt yet to push back against local opposition to the infrastructure underpinning its cloud and AI businesses. Garman framed the construction of new data centers as a race the United States cannot afford to lose.

"With any change, there will be important questions raised, but there will also be misinformation and outright lies, and in the age of social media and 24/7 news, myths take hold faster than ever before," Garman wrote. "In fact, this build out is s…"

Garman's intervention targets the three main fronts on which data center projects across the country have drawn resistance: concerns over jobs, power demands, and environmental impact. Rather than conceding ground on any of these, the AWS chief characterized much of the criticism as misinformation spreading faster than factual debate can correct it.

The blog post is a warning as much as an argument. Garman explicitly linked the pace of data center construction to AI dominance, saying "our nation can't afford to lose" that competition. The framing positions local permitting fights not as isolated zoning disputes but as inputs to a geopolitical contest — an argument designed to raise the perceived stakes for state and municipal officials weighing approval of Amazon's facilities.

For the semiconductor supply chain, the message matters because of what it signals about demand expectations. AWS remains one of the largest buyers of AI accelerators and server silicon, and the company's willingness to fight public battles over data center siting indicates Amazon expects the current buildout cycle to continue at scale. Cloud providers' capacity plans flow directly into orders for GPUs, custom AI chips, high-bandwidth memory, and advanced packaging.

The rhetorical escalation also fits a broader pattern among hyperscalers, who increasingly argue that grid constraints and community opposition — not chip supply — are becoming the binding limit on AI infrastructure expansion. Whether Garman's argument succeeds with the communities themselves will shape where, and how quickly, the next wave of AI compute capacity actually comes online.