Future steelmaking technologies and the gradual degradation of ore quality are forcing producers to change current mined iron ore workflows and product offerings from the Australian Pilbara region. Globally iron and steelmaking industries are working towards addressing the challenge of reducing greenhouse gas (GHG) emissions. There are many potential pathways to achieve lower emission steel production including hydrogen direct reduced iron with electric smelting furnace (DR-ESF), electrolysis, and other innovative technologies. Advancing the industry along any of the possible processing pathways via these process routes will require the assessment of the suitability of current ores, the design of new ore product streams and the extensive development of downstream steelmaking technology. OreAI aims to deliver a disruptive 4D product analysis workflow for iron ore to produce bespoke Pilbara iron ore products most suited to emerging supply chain requirements. OreAI builds proprietary 3D X-ray CT imaging techniques integrating them with bulk and 2D microanalysis data, and combining these with machine learning and artificial intelligence to create a novel workflow applicable across the mining value chain—from pit to furnace. An example of OreAI’s application for supporting orebody characterisation is described in the present study where 1200 lump samples of iron ore from a Pilbara asset are utilised. The results show that OreAI can be used to extend microanalytical information from 1D assay data and 2D microanalytical image data to 3D X-ray images of the ore material on a large batch of ore material across a range of material types. OreAI uses image analysis tools and AI to map and quantify porosity (connected and disconnected), textural information, mineral phase distributions and associations, density measures (apparent dry, wet, moist and skeletal) in 3D volumes. Tests of the model illustrate OreAI’s ability to accurately propagate high resolution 2D data to 3D and offers high quality mapping and an ability to define uncertainties in predictions. Detailed 3D distributions of mineralogy, textures and physical properties across material types can be derived for the orebody.