English
The global energy landscape in 2026 is defined by a striking paradox: a structural oversupply of liquefied natural gas (LNG) is set to slash European gas prices by nearly half by 2030, even as surging electricity demand from AI data centers pushes global grids to their limits. According to Wood Mackenzie forecasts released in early 2026, European traded gas prices could fall to an average of €24/MWh ($8/MMBtu) by 2030, down sharply from 2025 levels. Meanwhile, the International Energy Agency (IEA) projects that AI data center electricity consumption will reach 1,000 TWh by 2026 — equivalent to Japan's entire annual electricity usage — making power access the primary bottleneck for AI growth worldwide. This collision of cheap LNG and insatiable AI power demand is reshaping industrial competitiveness, energy security strategy, and investment flows across the US, Europe, and Asia.
The LNG Glut: A Supply Wave Like No Other
A wave of new LNG export capacity from the United States and Qatar is flooding global markets. The US Energy Information Administration (EIA) reports that US LNG exports are forecast to reach 17.0 Bcf/d in 2026, rising another 9% in 2027, driven by new projects including Cheniere's Corpus Christi Stage 3, Golden Pass LNG, Port Arthur LNG Phase 1, and Rio Grande LNG. Golden Pass LNG — a joint venture between QatarEnergy (70%) and ExxonMobil (30%) — shipped its first cargo in April 2026, adding 2.0 Bcf/d of nominal capacity. Total US LNG export capacity is on track to exceed 25 Bcf/d by 2030.
ABN AMRO's Energy Market Outlook 2026 forecasts TTF gas prices averaging €30/MWh in 2026, with summer prices falling to €26/MWh. The bank notes that new LNG capacity from the US, Canada, and Qatar should ease supply tightness and price volatility. Wood Mackenzie estimates that the LNG wave could cut European annual energy costs by roughly $46 billion by 2032, creating cumulative savings of nearly $213 billion. The European energy cost crisis that began with Russia's invasion of Ukraine may finally be reversing.
Why Prices Are Falling Despite Geopolitical Risks
Even with disruptions — including a March 2026 attack on Qatar's Ras Laffan facility that knocked out roughly 13 million tonnes of production capacity — the structural supply overhang remains intact. Rabobank analysts note that Gulf disruptions have temporarily tightened markets, pushing Q2 2026 TTF to around €61/MWh, but the medium-term trajectory remains bearish. Wood Mackenzie's base case sees European traded gas prices almost halving by 2030 compared to 2025 levels, trading at an average of €24/MWh in the 2030–2035 period. The global LNG market outlook is fundamentally oversupplied through the end of the decade.
AI's Power Hunger: 1,000 TWh and Rising
While LNG markets face a glut, electricity markets face a demand shock. The IEA's April 2026 report on Energy and AI projects that global data center electricity consumption will exceed 1,000 TWh by the end of 2026 — equivalent to Japan's total annual electricity use. AI-specific infrastructure is tripling over the same period. Morgan Stanley warns of 126 GW of additional global data center power demand through 2028, with a 49 GW generation shortfall looming in the US alone.
Bloom Energy's 2026 Data Center Power Report reveals that nearly one in three US data centers aim to go fully off-grid by 2030, using onsite solid oxide fuel cells (SOFC) to bypass grid constraints. Texas is projected to capture 30% of the US data center market by 2028 (up from 12% in 2023), while legacy hubs like California and Oregon could see their share drop by more than half. The report notes that 73% of operators are now embedding onsite power into their long-term strategies, as grid interconnection timelines stretch to 18–24 months versus 90 days for modular fuel cell deployment.
The Grid Crisis Behind the AI Boom
The power crunch is already causing real-world consequences. PJM capacity prices have spiked nearly tenfold, and utilities requested $31 billion in rate hikes in 2025 alone. Electricity costs have risen 42% since 2019, sparking a ratepayer revolt. In March 2026, seven major AI companies — including Amazon, Google, Meta, Microsoft, OpenAI, Oracle, and xAI — signed the White House Ratepayer Protection Pledge, committing to fund their own power infrastructure rather than passing costs to residential customers. The AI data center power crisis is now a central issue for grid regulators and policymakers.
The Collision: Cheap Gas Meets Insatiable Power Demand
The intersection of LNG oversupply and AI power hunger creates a complex dynamic. Cheap natural gas provides a ready fuel source for the gas turbines that are increasingly powering AI data centers. The IEA reports that natural gas currently supplies about 26% of global data center electricity, and in the US, that share exceeds 40%. Gas turbine orders surged 70% in 2025 as developers turned to onsite natural gas generation to bypass grid bottlenecks.
However, the same LNG glut that lowers European gas prices also threatens to raise US gas prices. Wood Mackenzie warns that US Henry Hub prices could rise nearly 50% above 2025 levels, averaging $4.90/MMBtu, as LNG exports and data center demand tighten the domestic market. This narrowing of the transatlantic price gap could erode the US industrial competitiveness advantage that was built on cheap energy. The US industrial competitiveness and energy costs are now under scrutiny as the energy landscape shifts.
Investment Implications: Winners and Losers
For Europe, cheaper gas is a lifeline for energy-intensive industries. Wood Mackenzie estimates that annual energy cost savings of €39 billion by 2032 could revive sectors like petrochemicals, metals, and iron and steel that have struggled since 2021. However, the EU's ambitious decarbonization agenda — with carbon prices above €80/tonne — means that cheap gas alone may not restore competitiveness. The EU carbon market and industrial policy will need to balance emissions reduction with industrial revival.
For the US, the LNG export boom creates a double-edged sword: export revenues and geopolitical leverage come at the cost of higher domestic energy prices. For Asia, lower LNG prices could accelerate the shift away from coal, particularly in price-sensitive emerging markets. The IEA projects Asian LNG demand will rebound by 14 million tonnes (+5%) in 2026 after a contraction in 2025, driven by lower spot prices.
FAQ: The 2026 Energy Paradox
What is the LNG glut of 2026?
The LNG glut refers to a structural oversupply of liquefied natural gas driven by massive new export capacity from the US and Qatar. Wood Mackenzie forecasts that European traded gas prices could nearly halve by 2030 compared to 2025 levels, falling to an average of €24/MWh.
How much electricity do AI data centers consume in 2026?
The IEA projects that global data center electricity consumption will exceed 1,000 TWh by the end of 2026, equivalent to Japan's entire annual electricity usage. AI-specific infrastructure is tripling over the same period.
Why is cheap LNG not solving the AI power problem?
While cheap LNG provides fuel for gas-fired power plants, the bottleneck is grid interconnection capacity and transmission infrastructure, not fuel supply. Data centers face 18–24 month grid hookup timelines, driving a shift toward onsite power generation.
Which regions benefit most from the LNG glut?
Europe is the biggest beneficiary, with potential annual energy cost savings of €39 billion by 2032. Asia also benefits from lower spot LNG prices, while the US faces higher domestic gas prices due to export demand.
Will AI data centers accelerate the energy transition?
In the near term, AI data centers are increasing reliance on natural gas and coal. However, the IEA projects that renewables will meet nearly 50% of new data center demand by 2030, and small modular reactors (SMRs) could enter the mix after 2030.
Conclusion: A Defining Energy-Economic Story
The collision of LNG oversupply and AI power demand is the defining energy-economic story of 2026. For the first time in years, Europe faces the prospect of genuinely cheap gas — but the infrastructure to convert that gas into affordable electricity for AI data centers is strained to breaking point. The winners will be those who can bridge the gap between abundant fuel and constrained power delivery. As Wood Mackenzie's analysis makes clear, the next decade will be shaped not by energy scarcity, but by the paradox of simultaneous abundance and constraint.
Sources
- Wood Mackenzie, 'Gas & LNG 2026 Outlook' and 'Energy Costs Across European Industrial Sectors Could Fall by €39 Billion by 2032' (2026)
- ABN AMRO, 'Energy Market Outlook 2026' (2026)
- IEA, 'Energy and AI: Energy Supply for AI' (April 2026)
- Bloom Energy, '2026 Data Center Power Report' (2026)
- US Energy Information Administration, 'Today in Energy' (2026)
- Morgan Stanley, 'AI Data Center Power Demand Forecast' (2026)
Follow Discussion