In Castro et al. (2024) electricity demand by datacentres is projected to consume all of the world’s electricity supply by 2033. Given the visibility of the Journal of Energy Policy it is necessary to clearly point out that this claim is based on flawed assumptions and poor methodological choices—most notably, the use of outdated and inappropriate data and a fundamental mischaracterisation of datacentre energy dynamics. The resulting forecast of electricity demand is inconsistent with existing literature by orders of magnitude. The ongoing growth in datacentre energy consumption requires informed and constructive discussion. The highly overstated findings in Castro et al. (2024) instead risk being a significant disservice to the public and policy debate.