TodayThursday, November 19, 2026

Years-Long Turbine Backlogs Signal Where The Real AI Infrastructure Trade Is Heading

The scramble to power artificial intelligence data centers has created shortages that go far beyond semiconductors, reaching deep into the natural gas supply chain.

Demand for natural gas turbines made by GE Vernova (GEV) has grown so intense that delivery backlogs now stretch for years, as data center developers race to secure their own electricity generation.

Of the 26 analysts who track GEV, the median one-year price target sits at $25, implying a potential gain of 28.6% from current levels, according to CNN-tracked data.

That bullish consensus reflects a broader conviction forming across Wall Street: the AI buildout is, at its core, an energy story, and natural gas sits at the center of it.

The thesis gained fresh momentum when Anthropic published its S-1 filing, mentioning natural gas five times within a single paragraph of its risk disclosure section.

One analyst summarized the sentiment bluntly in a note to clients: “We know power is a constraint to AI and Anthropic’s S-1 filing did nothing to assuage that view.”

The same analyst added: “AI compute is a natural gas demand story,” a framing that is rapidly becoming consensus among investors watching the sector.

Anthropic’s filing warned that as more data center operators shift toward natural gas-fired generation, regulatory friction could intensify competition for limited supply, stating the company expects “increased competition among users for the same natural gas supply and pipeline capacity in those areas.”

The filing also flagged constraints on pipeline capacity and firm fuel transportation as compounding risks, highlighting that securing the fuel itself is only part of the challenge.

To work around these limitations, data center developers have increasingly pursued what the industry calls “behind the meter” power, building their own generation infrastructure or purchasing capacity directly from an existing energy provider.

Natural gas is particularly attractive for this purpose because it remains cheap and abundant across the United States, giving it a practical edge over alternatives that require longer permitting timelines or greater infrastructure investment.

Investors have several ways to position for this shift, ranging from equity stakes in natural gas producers to shares of pipeline operators that transport fuel directly to data center campuses.

Unlike semiconductor bets, which hinge on which AI models succeed, natural gas companies stand to benefit regardless of which frontier lab ultimately dominates the market.

The structural demand for power will persist as long as data centers keep expanding, making energy infrastructure one of the more durable long-term plays attached to the AI cycle.

Raul Martinez

Raul Martinez covers crypto, AI, tech and iGaming news for iBusiness.News. He is especially interested in generative AI, robotics, and blockchain startups.