资料中心开发商正逐渐依赖客制化的天然气发电厂，以满足人工智慧热潮带来的庞大电力需求，并避开面临长时间延宕的电网连接问题。这项转变对气候目标构成了重大威胁，因为 99 座拟议中的电厂每年可能会排放约 3.18 亿公吨的二氧化碳，可能使美国电力部门的碳排放量增加高达 20%。

电力需求的激增主要由亚马逊和微软等大型科技公司，以及 OpenAI 和 Anthropic 等人工智慧实验室所推动。这些计划遍布 22 个州，其中德州因其丰富的石化燃料资源和宽松的监管环境而高度集中，促成了大规模的开发案，例如亚马逊占地 8,000 英亩的厂区，以及由雪佛龙为微软建造、为占地 2,000 英亩的资料中心供电的天然气厂。

这种对石化燃料的依赖直接与大型科技公司在人工智慧浪潮爆发前所做出的雄心勃勃的净零气候承诺相抵触。尽管这些公司坚称其气候目标并未改变，但批评者指出，他们早期的永续发展领导地位与目前快速部署高碳排新基础设施的行为之间存在著强烈的对比。

Data center developers are increasingly relying on bespoke natural gas power plants to meet the massive electricity demands of the artificial intelligence boom, bypassing grid connections that face lengthy delays. This shift poses a major threat to climate goals, as 99 proposed plants could emit about 318 million metric tons of carbon dioxide annually, potentially increasing US power sector emissions by up to 20%.

The surge in power demand is primarily driven by major tech companies like Amazon and Microsoft, alongside AI labs such as OpenAI and Anthropic. These projects span across 22 states, with a high concentration in Texas due to its abundant fossil fuel resources and relaxed regulatory environment, leading to massive developments like Amazon's 8,000-acre site and a gas plant built by Chevron for Microsoft's 2,000-acre data center complex.

This reliance on fossil fuels directly contradicts the ambitious net-zero climate pledges that Big Tech companies made prior to the artificial intelligence boom. While these companies maintain that their climate goals remain unchanged, critics point out a stark contrast between their earlier leadership in sustainability and their current rapid deployment of new, carbon-intensive infrastructure.