My story with this article goes back to February of this year. I opened ERR, saw the headlines regarding Iran, and thought, “Great, here we go again. Energy prices are going to skyrocket!” – and then they did. Despite that, the NASDAQ Composite kept rising until June, and has since stalled, down roughly 4% since June as of 22 October. This left me with one question: will AI’s promised gains outweigh the enormous cost of delivering them? Energy may prove to have the answer.
Energy and AI
AI runs on energy-intensive data centers, semiconductor supply chains and reliable electricity networks, all requiring substantial capital investment. The International Energy Agency (IEA) estimates that major technology companies spent over $400 billion on capital expenditure in 2025, with further growth in 2026. One example is Alphabet’s planned investment in Finland, with infrastructure expansion valued at least €13 billion. Capital expenditure of this scale can push up local electricity prices: Bloomberg found in 2025 that wholesale electricity in some US markets near data centers cost up to 267% more for a single month than 5 years ago, a spike recorded even before the outbreak of the Iran conflict.
The IEA projects that global data center electricity consumption will rise from 485 TWh in 2025 to 950 TWh in 2030. By then, it would account for around 3% of global electricity demand, with the projected total comparable to Japan’s current annual electricity consumption. The concern is not that AI will run out of power, but that infrastructure bottlenecks and higher operating costs weaken expected returns, in turn putting downward pressure on the share prices of technology companies involved in the buildout.
Energy in the economy
Energy is the lifeblood of the modern economy. When prices surge, businesses face higher costs while consumers lose purchasing power. The European Central Bank found that rising energy prices can force households to reduce spending or draw down savings. As demonstrated during the 2022 European energy crisis, such supply shocks heavily depress consumption. Today, renewed geopolitical tensions raise the risk of another prolonged energy shock. With the Middle East accounting for approximately 30% of global oil production, disruptions in the region have consequences far beyond its borders.
These risks are reflected in Goldman Sachs’ decision to raise its December 2026 Brent crude price forecast from $85 to $90 per barrel on 16 October 2026, citing heightened risks associated with the Iran conflict. Although the bank expects Middle Eastern supply to adapt over time, causing prices to fall, renewed escalation could keep them higher for longer. The IMF’s April 2026 outlook, meanwhile, warned that a prolonged or broader military conflict in the Middle East could significantly weaken growth while increasing inflation. In the outlook’s severe scenario, global growth would fall to around 2% – a level consistent with near-recession conditions. The effects are already visible in the travel industry, with airBaltic cancelling flights. If energy inflation persists alongside elevated borrowing costs, corporate margins and household spending could come under further pressure, making it harder for AI-driven productivity gains to offset broader economic weakness.
However, the largest AI companies have a cushion. Alphabet for instance operates with $55.9 billion in cash and equivalents, about $242 billion including short-term investments at the end of June 2026. That makes a collapse less likely, but it doesn’t remove the pressure on valuations if returns keep weakening.
Why it matters
Brent crude oil currently sits at nearly twice its price at the start of 2026, and is likely to stay elevated for longer. Interest rates globally are rising, with US bond yields above 5%, at 24-year highs.
Primary logistics routes for energy are also under pressure. The Strait of Hormuz, which carried about 20% of globally traded oil, was closed for over three months this year and remains exposed to renewed escalation, while the Bab al-Mandab strait is now nearly under Houthi control as well. All of these factors combined put cost-cutting pressure on governments and corporations, as well as regular people.
It is also important to note the concentration in today’s market. NVIDIA alone accounted for roughly 8% of the S&P 500 index, with the Magnificent 7 making up 34.9% of it according to The Motley Fool. Owning index funds is generally considered diversification, but the share of tech in the funds one owns is much larger than they might think. So a fall in AI stocks hits even a ‘diversified’ portfolio, which is why a cash reserve and spreading beyond US tech matter. A lesson that many, including myself, have found to be costly.
The needle?
Energy may not be the one needle that bursts the AI bubble, but it is one of the forces pushing it to the brink. Seeing the constant “deal, no deal” back and forth in the US-Iran negotiations, I exited most of my riskier positions during the summer. If I were to give one recommendation to someone beginning to invest today, it would be worth watching the Russell 2000 index, which tracks 2000 of the smaller, more economically vulnerable companies on the US stock market. In a near-recession environment, the smaller, riskiest companies are generally exited first.
Legal disclaimer: This is not financial advice.
Sources
1. International Energy Agency. Key questions on energy and AI. https://www.iea.org/reports/key-questions-on-energy-and-ai 2. International Energy Agency. Japan: how is electricity used? https://www.iea.org/countries/japan/electricity#how-is-electricity-used-in-japan 3. Google. Google’s AI commitment to Finland. https://blog.google/innovation-and-ai/infrastructure-and-cloud/global-network/google-ai-commitment-to-finland/ 4. European Central Bank (2022). Economic Bulletin article on energy prices and consumption. https://www.ecb.europa.eu/press/economic-bulletin/articles/2022/html/ecb.ebart202203_01~f7466627b4.en.html 5. Bloomberg (2025). AI data centers and electricity prices. https://www.bloomberg.com/graphics/2025-ai-data-centers-electricity-prices/ 6. International Energy Agency (2025). World Energy Investment 2025: Middle East. https://www.iea.org/reports/world-energy-investment-2025/middle-east 7. Financial Times. Article on energy supply routes. https://www.ft.com/content/993940f2-c9a3-4e20-8c0a-f73d448f090d?syn-25a6b1a6=1 8. ERR. airBaltic sulgeb järgmisel suvel Tallinnast 11 lennuliini; ERR News. airBaltic cancels Tallinn–London Gatwick flights for winter season. https://www.err.ee/1610125372/air-baltic-sulgeb-jargmisel-suvel-tallinnast-11-lennuliini https://news.err.ee/1610104081/airbaltic-cancels-tallinn-london-gatwick-flights-for-winter-season 9. Goldman Sachs. Will oil prices decline as Middle East supply recovers? https://www.goldmansachs.com/insights/articles/will-oil-prices-decline-as-middle-east-supply-recovers 10. International Monetary Fund (2026). World Economic Outlook, April 2026. https://www.imf.org/en/publications/weo/issues/2026/04/14/world-economic-outlook-april-2026 11. Nasdaq. Nasdaq Composite Index (COMP). https://www.nasdaq.com/market-activity/index/comp 12. Alphabet. Investor relations: earnings. https://abc.xyz/investor/earnings/ 13. Yahoo Finance. Brent crude oil futures (BZ=F). https://finance.yahoo.com/quote/BZ=F/ 14. Bloomberg. US government bond yields. https://www.bloomberg.com/markets/rates-bonds/government-bonds/us 15. The Motley Fool. The Magnificent Seven’s Market Cap vs. the S&P 500. 16. Google Finance. Russell 2000 compared with Nasdaq Composite and S&P 500, six-month view. https://www.google.com/finance/quote/RUT:INDEXRUSSELL?comparison=.IXIC%3AINDEXNASDAQ%2C.INX%3AINDEXSP&type=line&window=6M
← All articlesKõik artiklid