Cleveland-Cliffs has signed a three-year deal with Palantir to bring AI deeper into steelmaking, production planning, order entry, and daily operations across its manufacturing network.
Cleveland-Cliffs is taking Palantir from pilot project to core operating system. The steelmaker announced on April 28, 2026, that it had entered a three-year strategic partnership with Palantir Technologies to deploy advanced AI tools across its manufacturing footprint. The agreement puts Palantir's platform into production planning, order entry, and operational workflows, where small delays or bad assumptions can ripple through plants, customers, and supply chains. Chairman, President, and CEO Lourenco Goncalves said the pilot made the decision clear, arguing that integrated steelmaking is too complex for people and legacy systems to manage alone at full scale.
That complexity matters. Steel production depends on raw material availability, furnace utilization, customer specifications, transportation timing, energy costs, and maintenance windows that rarely move in perfect sync. A delay in one part of the chain can affect order fulfillment somewhere else. Palantir's ontology-based approach is designed to connect those moving parts into a usable operating model, giving managers a clearer view of constraints before they become expensive problems. According to Cleveland-Cliffs' own announcement, the goal is to better integrate data, anticipate bottlenecks, and coordinate activity across facilities in real time.
The deal also fits a broader moment for heavy industry. Bloomberg noted that Cliffs is stepping up modernization efforts as manufacturers face uneven demand and persistent cost pressure. Software has already remade logistics, retail, advertising, and finance, but steel remains a harder test. The equipment is physical, the margins are cyclical, and the decisions are tied to plants that cannot simply be restarted like a cloud server. That is why a serious AI deployment in steel is more than a technology headline. It is a test of whether enterprise AI can improve the rhythm of asset-heavy businesses.
Palantir has been building toward this kind of industrial work for years. Its Foundry and AIP products are meant to pull data from enterprise systems, sensors, planning tools, and supply chains into a shared decision layer. In a steel business, that can mean linking sales commitments to production schedules, linking maintenance needs to output targets, and giving operators a better view of how a change in one facility affects the rest of the network. The promise is not a chatbot sitting on top of a spreadsheet. It is a system that helps people act faster because the operational picture is less fragmented.
Other Palantir wins point in the same direction. Sumitomo Corporation has used AIP in oil country tubular goods operations, while Palantir has also worked with partners such as Rackspace to make Foundry and AIP easier to run in production environments. Those examples matter because enterprise AI is moving past demos. Companies are no longer asking only whether a model can generate an answer. They are asking whether it can sit inside real processes, respect controls, and help teams make decisions where mistakes carry financial consequences.
For manufacturers, the biggest gain is visibility. Enterprise resource planning systems, plant sensors, procurement tools, customer orders, and maintenance records often contain useful data, but they do not always speak to each other in time for managers to act. A platform that connects those sources can support decisions about scheduling, inventory, maintenance, and procurement. In a business like Cleveland-Cliffs, which is vertically integrated from iron ore and pellets through steelmaking and finishing, that visibility can be especially valuable.
Why Steel Now
Steel is an obvious candidate for operational AI because the industry is under pressure from several directions at once. Demand can swing with automotive production, construction activity, infrastructure spending, and energy markets. Input costs can move quickly. Customers want reliability, but mills have to balance that reliability against capacity, labor, raw materials, and equipment constraints. Data-driven systems cannot remove cyclicality from the steel market, but they can help a producer respond with less waste and fewer surprises.
Cleveland-Cliffs also has strategic reasons to sharpen its operating model. The company is a major supplier of automotive-grade steel in the United States and has exposure to higher-value areas such as electrical steel and motor laminations, where quality and timing matter. Better planning can protect margins when the market is soft and improve throughput when demand strengthens. The bull case is straightforward: if Cliffs can use AI to squeeze more productivity out of its existing footprint, the impact does not require a dramatic change in the steel cycle to matter.
Enterprise Implications
The larger implication is that AI adoption is shifting from office productivity to operational infrastructure. Early enterprise AI discussions focused heavily on writing, search, coding, and customer support. Those use cases are still important, but the next contest is inside the operating core of companies that move materials, run plants, manage fleets, or coordinate thousands of physical assets. Palantir wants to be the platform those companies use when generic software is not close enough to the work.
For Cleveland-Cliffs, the partnership will be judged by execution, not language. The company did not disclose financial terms, and it has not yet provided detailed return targets for the rollout. Investors and customers will want to see whether the system improves planning accuracy, reduces bottlenecks, supports better delivery performance, or helps lift margins over time. A three-year agreement gives both sides room to prove that the pilot can become a durable production tool.
Watch Forward
The next signals to watch are deployment pace and measurable operating results. If Cleveland-Cliffs can show that Palantir's platform improves coordination across plants, the deal will strengthen the case for AI in industries where complexity has long limited software's impact. It would also give Palantir another reference point outside the usual government and data analytics conversation. For steelmakers, the lesson is practical: competitive advantage may increasingly come from how well a company connects its operations before the market turns again.
Also read: AI-powered Russian cyber campaigns are turning Poland into Europe's most attacked country • Spotify's AI music problem is bigger than a filter button • Venture capital hires its first chief AI officer and the entire industry must catch up
