UK gas use is expected to surge sharply as data centres expand rapidly, the Financial Times reported on Nov 6, 2026. Grid connection delays are pushing data centre developers to invest in fossil fuel generation to secure power supply, the report said.

On the same day, a Stack Overflow blog interview discussed the same issue from a broader perspective. Rakesh Kumar, professor of computer engineering at the University of Illinois Urbana-Champaign, spoke about the complexity of modern semiconductor manufacturing, the environmental and geopolitical footprint of AI data centres, and how AI models are changing the future of chip design.

Grid bottlenecks and the pivot to fossil fuels

The FT report links the gas surge directly to grid connection delays. Because waiting times to connect data centres to the public grid are too long, some developers are choosing to invest in their own fossil fuel generation to meet the power needs of computing infrastructure. The trend means AI compute expansion is hitting real energy bottlenecks and could carry an extra carbon cost.

The report did not disclose the specific scale of the gas increase, the number of affected data centre projects or the names of the developers involved. The delays may be related to tight UK grid capacity and lengthy approval processes, but the FT report did not elaborate.

Academic view: the massive footprint of data centres

In the Stack Overflow interview, Kumar provided context from the perspective of chip manufacturing and AI infrastructure. He noted that manufacturing modern semiconductors is extremely complex, and that AI data centres have a massive environmental and geopolitical footprint. The interview also covered how AI models are rapidly shifting the future of chip design, suggesting that the interaction between compute demand and chip supply is reshaping the entire supply chain.

Ryan sits down with Rakesh Kumar, professor of computer engineering at the University of Illinois Urbana-Champaign, about the astounding complexity of manufacturing modern semiconductors, the massive environmental and geopolitical footprint of AI data centers, and how AI models are rapidly shifting the future of chip design.

The interview summary did not include direct comment from Kumar on UK gas use or grid connection problems, but his remarks on the environmental footprint of data centres echo the energy and carbon issues raised in the FT report.

What to watch

  • The exact scale of growth in UK data centre gas use and its impact on carbon targets;
  • The causes of grid connection delays and regulatory responses;
  • The scale, type and sustainability of developer investment in fossil fuel generation;
  • The long-term interaction between AI compute expansion, chip manufacturing and energy infrastructure.

Together, the two threads — a market report and an academic interview — sketch the outline of the energy bottlenecks and carbon costs behind AI compute expansion. The FT report provides specific developments in the UK market, while the Stack Overflow interview adds macro context on semiconductor manufacturing and the environmental footprint of data centres. Both point to an emerging issue: the energy demand of computing infrastructure is becoming an unavoidable constraint for the tech industry.