A preliminary bill approved by a large majority in the US Senate could pave the way for more gas-fired electricity generation at artificial intelligence data centres, according to analysis published by OilPrice.com. The outlet said the Bipartisan Affordability and Jobs Act, known as BAAJA, would support new generating facilities and associated infrastructure, potentially shifting the balance of power between technology companies and electric utilities.

The analysis argues that continued rapid expansion of data centres requires faster construction of electricity transmission lines and gas pipelines, fewer opportunities for legal challenges and the relaxation of environmental constraints. OilPrice.com interprets the draft as addressing all three, including by limiting the ability of state attorneys general and tribal authorities to block infrastructure. However, it cautions that the legislation remains at an early stage, could undergo extensive revisions and might fail in a newly constituted House of Representatives.

Gas combustion turbines are central to the outlet’s assessment of future data-centre electricity supply. It reported that Anthropic had announced a multibillion-dollar purchase of equipment, mostly gas combustion turbines, for a new Texas facility. The analysis envisages a possible 40-50 gigawatts of additional gas-fired capacity at proposed facilities, but questions the wider grid benefits, arguing that these turbines would generate electricity at uncompetitive costs for much of the year.

The outlook for cheaper electricity is more promising in Texas, the analysis suggests, citing a September grid-connection update from the Electric Reliability Council of Texas. Requests covered 158 GW of solar capacity, compared with 38 GW of utility-scale solar already installed. A further 166 GW of batteries and 49 GW of wind capacity were also seeking connections.

Those figures do not represent confirmed construction: OilPrice.com noted that some projects would not secure financing or be built, while duplicate requests inflated parts of the queue. Nevertheless, the analysis said its scale pointed to the possibility of cheaper electricity for Texas consumers, although that outcome could come at the cost of future utility profitability.