ArcelorMittal partners with AWS to apply AI to lower-carbon steelmaking

ArcelorMittal is working with Amazon Web Services, the cloud-computing business of Amazon, on an AI and cloud initiative aimed at improving how the steelmaker approaches lower-carbon production, according to a report published by Yahoo Finance.

The collaboration centers on bringing AWS technology to ArcelorMittal’s operations and planning processes, with the goal of using advanced analytics and machine learning, a form of AI that learns patterns from data, to support decisions in manufacturing and supply. The report frames the effort as part of ArcelorMittal’s broader push to expand the availability of lower-carbon steel used in construction.

Both companies characterized the work as spanning multiple years, suggesting the project is intended to mature beyond pilots. However, the report did not provide publicly available implementation specifics such as which data sources ArcelorMittal would connect to the cloud, how the AI models would be validated on production lines, or what measurable performance targets were agreed up front.

While the report is focused on AI and cloud enablement, the business rationale is familiar in heavy industry: emissions reductions typically depend on a complex chain of inputs, including energy sourcing, process parameters, and logistics. Cloud platforms can centralize operational data, and machine learning can potentially help identify process settings that reduce emissions or material waste. In that context, the companies’ emphasis on “AI in lower-carbon construction” points to applications that may touch both production planning and downstream product requirements.

AWS has positioned itself for years around enterprise adoption of cloud and AI services, including data management, analytics, and machine learning tooling. Amazon’s corporate communications describe AWS as the company’s platform for cloud services used by organizations across industries, underscoring why a steelmaker would look to AWS for the underlying infrastructure and AI development environment.

A key limitation from the public reporting is what was not disclosed. Yahoo Finance’s account did not spell out the contract’s financial terms, the geographic scope of deployment, whether the system would be used by plant operators in real time or by analysts in planning workflows, or how quickly results would be measured against emissions or quality benchmarks.

For Amazon and AWS, deals like this are another example of how cloud providers increasingly court large industrial customers beyond traditional IT workloads, particularly where data availability and operational complexity can make AI initiatives difficult to stand up quickly. For ArcelorMittal, partnering with a hyperscaler also indicates a desire to industrialize analytics and AI capabilities rather than treating them as purely experimental projects.