Americans’ electricity bills are up. Don’t blame AI

Were it not for data centres, prices might be even higher

Photograph: Shutterstock

Listen to this story

Your browser does not support the <audio> element.

P OWER BILLS are going up in America and the people are angry. They know whom to blame—the bosses of technology firms thirsting for more juice to fuel artificial-intelligence data centres. Ashburn, a town of 45,000 in a featureless part of Virginia that has earned the nickname “Data Centre Alley”, has some 150 of these. They consume roughly as much electricity as Philadelphia, a city of 1.6m. On March 4th Donald Trump convened tech leaders to sign a pledge to “build, bring or buy their own power supply…ensuring that Americans’ electricity bills will not increase”.

Their solemn pledges notwithstanding, the chief executives can do little to contain prices. That is not, though, because AI is unstoppable. It is because the AI boom is not chiefly to blame for the rising costs.

In the past few years retail electricity prices have indeed outpaced overall inflation (see chart 1). And data centres are gobbling up more power. Goldman Sachs, a bank, reckons that they will account for nearly half of the overall demand growth in America in the coming years.

Yet even bullish forecasts put data centres’ share of total demand at only a fifth in 2030. Today it is less than a tenth. A study last year by the Lawrence Berkeley National Laboratory showed that data-centre load was not the main cause of the rate rises in the five years to 2024. It fingered grid upgrades and rising costs of power-generating equipment and raw materials such as copper. Wood Mackenzie, a research firm, estimates that last year demand for distribution transformers outstripped supply by 10%. For power transformers the gap was 30%. Manufacturers report waiting lists for essential grid-related kit stretching to 120 weeks or more, up from 50 weeks in 2021.

Chart: The Economist

Many prices started going up in early 2021, nearly two years before the launch of Chat GPT ignited the AI boom. They are likely to keep rising for non- AI reasons. The Edison Electric Institute, which represents private-sector utilities, predicts its members’ cumulative capital spending will reach $1.1trnbetween2025and2029,upfrom$765bn in the previous five years. More than half the sum for distribution and transmission infrastructure will go on replacing ageing equipment and hardening it against extreme weather made likelier by climate change. Between 2019 and 2023 big Californian utilities spent $27bn just on mitigating wildfire risk. These investments have been neglected for years. Now, says an industry bigwig, AI provides a pretext to help win approval from regulators to pass the cost on to consumers.

And these are not the only non- AI cost pressures. Even before the war in Iran caused natural-gas prices to rise, analysts were predicting that domestic buyers would be increasingly competing with foreign ones as more export terminals for liquefied natural gas come online. Mr Trump, an inveterate renewables sceptic, has not helped by impeding the growth of solar and wind capacity. Peter Fox-Penner of the Brattle Group, a consultancy, notes that as a result prices are rising needlessly for the cheapest forms of new power generation.

AI may even be lowering prices. The tech giants are already investing in their own capacity (mostly, whisper it, in the clean variety). Microsoft has signed a long-term deal to restart a nuclear reactor at Three Mile Island to supply its data centres. Meta has backed a handful of nuclear startups. In December Google’s corporate parent, Alphabet, paid $5bn for Intersect Power, a developer of utility-scale solar power and battery storage. A data centre in Ashburn belonging to Equinix, a big operator, is experimenting with fuel cells.

Besides adding its own supply, big tech is making existing capacity more flexible. Google has agreed to novel tariff arrangements with Indiana Michigan Power, a midwestern utility, whereby its data centres can reduce their consumption when other demand is high. Microsoft is going further. In one of its Irish data centres it uses backup batteries as a “grid stabiliser” that can push power back into the network or draw excess power from it at times of stress. Since grids often run well below full capacity, adding a large, flexible customer can bring in lots of revenue for utilities without requiring costly expansion. This lets the utilities lower rates for households while preserving their margins.

The Electric Power Research Institute, a think-tank, found that some states with high load growth between 2019 and 2024 reported price declines, after adjusting for inflation (see chart 2). The World Resources Institute, another think-tank, notes that in North Dakota rising demand from oil and gas extraction, cryptocurrency miners, data-centre operators and food-processors led to large price reductions for local electricity users. PG & E, a big Californian utility, estimates that adding a gigawatt of load could lower bills by up to 2%. If Americans want lower electricity bills, they should be shouting for more AI, not less. ■

---

논증 분석

유형: causal

핵심 주장

미국 전기요금 상승의 주범은 AI 데이터센터가 아니라 노후 인프라 교체, 기상이변 대응, 천연가스 가격 등 구조적 요인이며, AI는 오히려 전기요금을 낮출 수 있다.

논리구조

1. 전제: 미국 소비자들은 AI 데이터센터를 전기요금 상승의 주범으로 지목하고 있으며, Donald Trump은 테크 기업 CEO들에게 자체 전력 공급 확보를 촉구했다.
2. 진단: 최근 몇 년간 소매 전기요금이 전반적인 물가상승률을 초과했고, 데이터센터의 전력 소비는 증가하고 있다. Goldman Sachs는 데이터센터가 향후 미국 전체 수요 증가의 절반 가까이를 차지할 것으로 전망한다.
3. 반론: 그러나 Lawrence Berkeley National Laboratory의 연구에 따르면, 2024년까지 5년간의 요금 인상의 주원인은 데이터센터가 아니라 전력망 업그레이드 및 구리 등 원자재 가격 상승이었다. 데이터센터의 전체 전력 수요 비중은 현재 10% 미만이며, 2030년에도 20%에 불과할 전망이다.
4. 논거: 전기요금 상승은 ChatGPT 출시로 AI 붐이 시작되기 약 2년 전인 2021년 초부터 이미 시작되었으며, 노후 장비 교체·기후변화 대응·산불 위험 완화(캘리포니아 유틸리티만 270억 달러 지출) 등 비AI 요인이 주도하고 있다.
5. 논거: Edison Electric Institute는 2025~2029년 회원사들의 누적 자본 지출이 1조 1,000억 달러에 달할 것으로 예측하며, 이는 이전 5년(7,650억 달러) 대비 큰 폭으로 증가한 수치다. 업계 관계자는 AI가 오히려 규제 당국을 설득해 비용을 소비자에게 전가하는 구실로 활용되고 있다고 지적한다.
6. 논거: Donald Trump의 재생에너지 규제와 LNG 수출 터미널 확대로 인한 천연가스 가격 상승 압력 등 추가적인 비AI 비용 요인도 존재한다.
7. 처방: AI 빅테크 기업들은 오히려 자체 전력 공급을 확대하고 있다. Microsoft는 Three Mile Island 원전 재가동 계약을 체결했고, Meta는 핵 스타트업을 지원하며, Alphabet은 유틸리티 규모 태양광·배터리 개발사 Intersect Power를 50억 달러에 인수했다.
8. 처방: 빅테크는 기존 전력망의 유연성도 높이고 있다. Google은 Indiana Michigan Power와 피크 시간대 소비를 줄이는 특수 요금제를 체결했고, Microsoft는 아일랜드 데이터센터에서 백업 배터리를 ‘그리드 안정화 장치’로 활용해 전력망에 전기를 역공급하고 있다.
9. 논거: Electric Power Research Institute의 연구에 따르면 2019~2024년 수요가 크게 증가한 일부 주에서는 오히려 실질 전기요금이 하락했다. PG&E는 1기가와트의 부하를 추가하면 전기요금이 최대 2% 낮아질 수 있다고 추정한다.

결론

AI 데이터센터는 전기요금 상승의 주범이 아니며, 오히려 자체 전력 공급 투자와 그리드 유연성 확대를 통해 전기요금을 낮추는 데 기여할 수 있으므로, 미국인들은 AI를 반대할 것이 아니라 더 많은 AI 투자를 요구해야 한다.

For more expert analysis of the biggest stories in economics, finance and markets, sign up to Money Talks, our weekly subscriber-only newsletter.