The government announced on Sunday a £2 billion public-private investment in a new artificial intelligence supercomputer campus to be built on the outskirts of Manchester, in what ministers described as the largest single commitment to AI computing infrastructure in British history. The facility, expected to be operational by the end of 2027, will house computing clusters capable of training frontier AI models and is intended to give UK researchers, universities and companies access to computing power that currently requires them to rely on American cloud providers or travel abroad.
The announcement was made by Science and Technology Secretary Peter Kyle at a keynote address at the Manchester Science Festival’s spring edition, attended by technology executives, university leaders and representatives from the UK’s AI research community. Kyle framed the investment as a direct response to the global AI infrastructure race, in which the United States and China have made enormous commitments to building the computing capacity needed to develop and deploy the next generation of artificial intelligence systems. “We cannot afford to be spectators in the most consequential technological competition of our lifetimes,” Kyle said.
The Manchester AI Campus, as the project has been named, will be built on a 42-acre site near Manchester Airport and will house computing infrastructure operated by a new public body, UK AI Compute, established as a subsidiary of UK Research and Innovation. Private sector partners including Arm Holdings, DeepMind, Microsoft and a consortium of British technology companies will co-invest alongside the government, contributing both capital and technical expertise. Microsoft has committed to deploying 100,000 AI accelerator chips at the facility, making it one of the largest concentrations of AI computing capacity in Europe.
The facility will serve a dual purpose: providing subsidised computing time to UK universities and publicly funded research institutions working on AI for science, medicine and public services; and offering commercial computing capacity to British AI companies and startups unable to access the scale of infrastructure available to their American and Chinese competitors. A tiered pricing system will prioritise academic and public-benefit research at the lowest rates, with commercial users paying market rates that will partially subsidise the public-benefit allocation.
The announcement was broadly welcomed by the UK tech sector, though some executives argued that the government needed to move faster. Representatives from the Alan Turing Institute, the Francis Crick Institute and several leading UK AI research groups said the facility would meaningfully address the compute bottleneck that had been forcing British researchers to either scale back their ambitions or apply for access to American facilities — a process that could take months and was subject to export control restrictions.
Critics from the opposition pointed to what they described as a gap between the scale of UK AI investment and the commitments being made in comparable economies. The United States has committed $500 billion to AI infrastructure through the Stargate initiative, while the European Union has pledged €200 billion over the next decade. The UK’s £2 billion commitment, while significant domestically, represents a fraction of these figures — though ministers argued that the UK’s strength lies in research talent and institutional quality rather than raw computing scale.
— Sarah Mitchell, London Capital Post
