Total Energies AI Operating System, US$108 M Mistral AI Deal, 3-Year Cognite Deployment, and 5 Key Agreements (2025 to 2026)

Total Energies’ Shift From AI Pilots to an Enterprise-Wide Operating System

In 2025, Total Energies executed a decisive strategic shift, moving beyond isolated, high-impact artificial intelligence projects to the full-scale construction of a standardized, enterprise-wide AI operating system. This transition is defined by a series of major partnerships designed to create a federated technology stack, leveraging external specialists to build core capabilities rather than relying on slower, in-house development alone. The company’s goal is to embed AI as a core enabler for both optimizing legacy upstream assets and accelerating its expansion into new energy models.

Pre-2025: Isolated High-Impact AI Pilots

Prior to 2025, Total Energies’ AI efforts were characterized by successful but functionally siloed applications targeting specific operational problems. These initiatives proved the value of AI in discrete use cases, such as the system developed to predict electrical submersible pump failures with 95% accuracy, which directly reduced downtime and production losses. While effective, these projects were not part of a unified, scalable platform, representing successful proof-of-concept deployments rather than a cohesive, enterprise-level digital strategy.

Post-2025: A Federated, Enterprise-Scale AI Platform

Beginning in 2025, Total Energies’ strategy matured into the creation of a comprehensive AI ecosystem. This is anchored by multi-year, platform-level agreements with technology leaders like Cognite, Emerson, and Mistral AI. The three-year global deployment of Cognite Data Fusion® across all upstream assets marks a fundamental move to standardize data operations. This federated model, where different partners provide best-in-class components of the AI stack, is designed to accelerate innovation and ensure that AI capabilities are scalable and integrated across the entire business, from industrial processes to corporate functions.

$108 M for Mistral AI, Total Energies Cements Sovereign AI Strategy

Total Energies’ investments in 2025 and 2026 are focused on building proprietary and sovereign AI capabilities, moving beyond simple technology acquisition to active co-development. Major capital commitments to its Pangea 5 supercomputer and a frontier AI program with Mistral AI show a clear strategy to control its own technological destiny, particularly with data-intensive applications for exploration and engineering. These investments provide the foundational infrastructure required to train and deploy bespoke AI models at an industrial scale.

Total Energies’ €100 Million Supercomputer Investment

The commitment of over €100 million for the Pangea 5 supercomputer in May 2026 is a critical infrastructure investment that underpins the company’s entire AI ambition. By increasing its computing power sixfold, Total Energies created the capacity to process the vast datasets required for advanced subsurface imaging and, crucially, to train the next generation of large-scale AI models developed in partnership with firms like Mistral AI. This high-performance computing power is essential for running complex simulations that both optimize traditional oil and gas projects and model new renewable energy systems.

The US$108 Million Mistral AI Frontier Model Program

The US$108 million joint program with Mistral AI, announced in September 2026, represents a direct investment in creating proprietary, frontier AI models. The collaboration is explicitly aimed at processing petabytes of subsurface data to accelerate decision-making in oil and gas exploration. This partnership goes beyond a typical vendor relationship, positioning Total Energies as a co-developer of sovereign AI solutions tailored to the energy sector, giving it a potential competitive advantage in geological analysis and resource management.

Table: Total Energies Key AI-Related Investments (2025 – 2026)

Total Energies Partnership Model: 5 Deals Building Its AI Stack (2025-2026)

Total Energies’ 2025 AI strategy is defined by a “best-of-breed” partnership model that assembles a complete technology stack by engaging specialized leaders for different layers of its digital infrastructure. This federated approach allows the company to rapidly integrate robust, purpose-built solutions for foundational data collection, industrial AI platforms, generative AI development, and autonomous operations, avoiding the pitfalls of a single-vendor or monolithic in-house build.

The Foundational Data Layer: Emerson and Cognite

The company’s AI stack begins with a robust data foundation built through two key partnerships. In July 2025, Total Energies signed a strategic agreement with Emerson to deploy advanced technologies for real-time data collection across its industrial sites. This raw data then feeds into the industrial data and AI platform provided by Cognite, whose Cognite Data Fusion® product is being rolled out across all global upstream assets in a three-year program started in September 2025. Together, these partners provide the essential infrastructure for collecting, contextualizing, and serving up data for AI applications.

The Application Layer: Honeywell and Mistral AI

At the application layer, Total Energies is partnering to deploy and co-develop AI tools for specific use cases. The November 2025 pilot with Honeywell to use its Experion Operations Assistant at the Port Arthur refinery is a direct test of AI-assisted autonomous operations. In parallel, the June 2025 collaboration with Mistral AI focuses on the next frontier: co-developing generative AI platforms tailored to the company’s multi-energy strategy, including applications for its growing renewable energy systems.

Table: Total Energies Key AI Partnerships and Projects (2025)

Global Rollout: Total Energies Deploys AI From Texas to Global Upstream

Total Energies’ AI strategy is explicitly global in scope, shifting from regional or asset-specific pilots to standardized, worldwide deployments. Foundational agreements, particularly the one with Cognite, are designed for deployment across the company’s entire operated asset base. This global ambition is complemented by strategic pilot projects, like the one at its Port Arthur, Texas refinery, which serve as crucial testbeds for validating technologies before a wider scale-up.

Total Energies’ Global Upstream Asset Deployment

The most significant indicator of this global approach is the three-year commercial agreement with Cognite, finalized in September 2025. The project’s mandate is to deploy its industrial data and AI platform across all of Total Energies’ operated upstream assets globally. This initiative aims to standardize how data is managed from drilling to production, creating a consistent digital foundation that allows AI tools and workflows to be developed once and deployed everywhere, a marked change from the previous asset-by-asset approach.

Port Arthur, Texas: A Key Pilot Site for Autonomy

While the strategy is global, its execution relies on localized validation. The November 2025 agreement with Honeywell to pilot its AI-assisted Experion Operations Assistant at the Port Arthur, Texas refinery is a key example. This major US industrial site serves as a real-world environment to test and refine autonomous operations technology. A successful pilot here will provide a blueprint and a justification for rolling out similar AI-assisted control rooms to other complex facilities around the world.

From Predictive Analytics to Generative AI: Total Energies’ Technology Shift

In 2025, Total Energies’ AI adoption matured from established predictive maintenance applications to the strategic co-development of generative and agentic AI, signaling a significant shift from consuming off-the-shelf AI to actively creating proprietary AI capabilities. This progression demonstrates a sophisticated, multi-layered technology strategy that leverages mature, proven technologies for immediate operational gains while simultaneously investing in frontier technologies for long-term competitive advantage.

Mature Use Cases: Predictive Maintenance and Analytics

The foundation of the company’s AI success lies in its long-standing use of predictive analytics for operational efficiency. Use cases that were proven before 2025, such as using AI to predict pump failures or applying Trend Miner’s analytics software to anticipate heat exchanger fouling, continue to deliver value. These applications are mature, reliable, and provide the tangible cost savings and uptime improvements that justify broader investment in more advanced digital technologies.

Frontier Technologies: Generative and Agentic AI

The defining feature of the post-2025 strategy is the aggressive move into frontier AI. The collaborations with France’s Mistral AI to build custom generative AI models and with Microsoft to deploy AI-powered autonomous agents represent a forward-looking effort to build next-generation tools. Unlike predictive analytics, which analyzes past data to forecast future events, these initiatives aim to create AI that can generate novel solutions, automate complex workflows, and interact with operators in a more dynamic way, pushing the boundaries of what is possible in an industrial setting.

SWOT Analysis for Total Energies’ Federated AI Strategy

Total Energies’ federated AI strategy, which relies on a network of specialized partners, provides rapid access to best-in-class technology and de-risks development by avoiding a monolithic internal build. However, this approach also introduces dependencies on key partners and requires strong internal governance to ensure seamless integration between disparate systems. The strategy’s success hinges on balancing the speed of external innovation with the need for a cohesive, secure, and internally managed digital core.

Table: SWOT Analysis for Total Energies AI Initiatives

Total Energies Scenario: Validating the $5 B AI-Enabled Savings Target

The success of Total Energies’ entire AI strategy will be validated by its ability to deliver on the expanded $5 billion cash savings program targeted for 2026-2030, making the measured financial impact of the Cognite and Mistral AI partnerships the most critical signal to watch. If these platform-level initiatives fail to translate into quantifiable operational efficiencies and cost reductions at scale, the significant upfront investment in building a federated AI ecosystem will be difficult to justify.

The Cognite Scaling Signal

The most immediate indicator of success will be the progress of the three-year global deployment of the Cognite Data Fusion® platform. Watch for company reports or technical presentations that provide metrics on the number of assets integrated, the speed of data analysis, and, most importantly, specific examples of operational cost savings or production gains attributed directly to the platform’s use. A smooth, on-schedule rollout with clear ROI will validate the foundational layer of the AI strategy.

The Mistral AI Prototype Signal

The collaboration with Mistral AI is a longer-term, higher-risk, and higher-reward play. The key signal to watch will be the emergence of the first tangible prototypes or applications from this partnership, particularly any that are applied to optimizing renewable energy operations or that demonstrate a measurable improvement in exploration success rates. Public announcements of successful model development or pilot deployments will be the first validation that this investment in frontier AI can deliver practical, industrial value.

The questions your competitors are already asking

This report covers one angle of TotalEnergies’ AI strategy. The questions that matter most depend on your work.

• Shell AI partnerships
• European sovereign AI for energy sector
• Measuring ROI on industrial AI deployments
• Cognite Data Fusion use cases in oil and gas

This report does not answer these. Enki Brief Pro does.

Your question, your angle, your framework. SWOT, PESTL, scenario modelling. The same niche depth, built around the decision your work actually depends on.

Run your first brief in Enki Brief Pro