77% of organisations have delayed AI infrastructure expansion over energy and sustainability concerns

More than three-quarters of organisations have delayed or restructured plans to expand AI infrastructure because of sustainability or energy concerns, with 36% describing that restructuring as significant, according to Seagate Technology's inaugural Data Infrastructure Readiness Report, published on 14 September.

The research, conducted independently by Recon Analytics among 2,712 enterprise technology decision-makers across the United States, China, India, the United Kingdom, Germany, France and Japan, points to a caution that sits awkwardly alongside stated investment intent. Some 76% of organisations rank data centre investment among their top three infrastructure priorities and 20% treat it as their single highest priority, yet only 38% say they are fully prepared for the long-term data demands of AI. AI-driven energy consumption was the leading environmental concern at 52%, followed by carbon emissions from that consumption at 51%.

With expansion under scrutiny, enterprises are looking to extend what they already own. Some 97% of organisations agree that extending infrastructure lifecycles significantly improves sustainability, and 94% expect their storage operations to become more sustainable over the next five years. That shifts competitive advantage among suppliers towards density, durability and lifecycle economics rather than capacity volume alone.

Seagate describes its own response as Sustainable Scaling, which it defines as increasing AI capacity and business value while continuously improving the efficiency of the infrastructure supporting it. “AI is reshaping the way organisations plan, build and operate infrastructure,” said Melyssa Banda, senior vice president of Edge Storage Business at Seagate Technology. “As data volumes grow, so does the value organisations can derive from the data. They need data infrastructure that helps them preserve, access and use more of that data over time.”

The restraint is not for want of demand. Some 99% of organisations expect AI to increase their storage requirements over the next three years, and 32% expect those needs to grow by more than half. Data quality and readiness was cited as a leading challenge to AI deployment by 53% of respondents and storage infrastructure by 43%, both ahead of compute availability at 27% and energy constraints at 24%.

The pressure comes from success rather than experimentation. Some 86% of organisations report moderate or significant returns from AI investments, including 33% reporting significant measurable returns, and 98% agree AI is turning storage into strategic business infrastructure.

Little of the shortfall traces back to available hardware. The leading barriers to greater preparedness were AI strategy maturity at 16%, budget and resources at 14%, and data management and governance at 14%, none of which is resolved by adding capacity. For suppliers positioning themselves as infrastructure partners, that locates the bottleneck inside enterprise decision-making rather than in supply.

“The next phase of AI will require capacity growth, but capacity alone will not be enough,” Banda said. “It will be defined by smarter infrastructure decisions on how effectively organisations scale, organise, retain and use the data that AI depends on. The companies that create lasting value from AI will be the ones that treat data infrastructure as a business strategy.”

On the evidence of this survey, the constraint on the next wave of enterprise AI deployment sits further down the stack than most budget conversations have so far reached.