AI News/AI News

Data center emissions data puts pressure on AI-reliant marketing teams

A new study, Meta's retreat from a 2030 target and rising public backlash make the energy cost of AI harder for brands to ignore.

Key points

  • A Nature Sustainability study counted 403 US hyperscale data centers using satellite imagery and estimated they used about 82 TWh from May 2024 to April 2025, roughly 1.8% of US electricity.
  • About 54% of that power likely came from gas and coal, producing an estimated 25.7 million metric tons of CO2, comparable to Denmark's annual footprint.
  • Meta's 2026 sustainability report says it is re-evaluating its 2030 net-zero timeline after total emissions rose from 15.6 to more than 21.7 million metric tonnes.
  • Backlash over data centers is spreading; TechCrunch ties it to the recent midterm results, and Microsoft, Amazon and OpenAI have talked about dropping NDAs with local governments.

Why it matters for marketers: AI-heavy marketing stacks sit on power-hungry infrastructure, so sustainability claims, vendor choices and brand reputation will face closer scrutiny from regulators, communities and customers.

A new paper in Nature Sustainability estimates that 403 hyperscale data centers in the US consumed around 82 terawatt-hours of electricity between May 2024 and April 2025. That is about 1.8 percent of national consumption. Because roughly 54 percent of the power likely came from gas and coal, the researchers put the resulting emissions at 25.7 million metric tons of CO2. They worked from satellite imagery and electrical capacity because operators do not disclose usage.

The same week, Meta's 2026 sustainability report said the company is re-evaluating its 2030 net-zero timeline and near-term intensity targets. Its total emissions rose from 15.6 million metric tonnes in 2024 to more than 21.7 million in 2025, which Meta tied largely to capital goods such as servers. Ars Technica notes that Amazon, Google and Microsoft have kept their original goals so far, though all face the strain of rapid buildouts.

For marketers, the issue is no longer abstract. Teams are adopting AI tools for content, personalization and automation, and clients and customers increasingly ask what that costs. The study also shows location matters: emissions per kilowatt-hour in Wyoming were roughly seven times those in Oregon. Meanwhile, TechCrunch reports that data center opposition is bipartisan and was a motivator in recent midterm results.

Watch for a bipartisan bill in Congress that could require national disclosure of data center electricity use, and for the researchers' planned site tracking individual facilities. Brands making sustainability claims around AI-assisted work should ask vendors for energy and emissions data and avoid overstating what they can verify.

Sources

This brief is our own summary. Read the full reporting at the source.

You read about AI automation. We run it.

DocFluence, built by Z Media, writes, designs, films and publishes for your brand every day, in your voice, from one login.