PharosAI, a U.K. research consortium, will use 10x Genomics’ Xenium spatial platform to build multiomic, spatially resolved cancer datasets from decades of archived NHS tumor samples. Supported by £18.9 million from the UK Research Ventures Catalyst, the collaboration will integrate genomic, transcriptomic and imaging data with AI models and intends to make datasets broadly available to accelerate diagnostics and drug discovery. The effort underscores growing public‑private initiatives to unlock hospital biobank value for translational AI workflows.