TotalEnergies and Mistral Plan More Than €100 Million AI Push for Oil and Gas Reservoirs

TotalEnergies and Mistral are putting more than €100 million behind a three-year effort to build AI models for oil and gas reservoir exploration, characterization, and development. The program’s central bet is that AI can turn vast stores of geoscientific data into more useful options for human experts deciding where and how to develop reservoirs.

The companies said the models will be designed to absorb, process, and interpret large volumes of subsurface information. Their intended output is not a single answer: the tools are meant to generate multiple development scenarios for exploration opportunities and help optimize or extend the life of existing projects.

From raw subsurface data to competing options

The companies say the program will use agentic AI to combine nearly 10 petaflops of data with almost a century of TotalEnergies’ geoscience and reservoir-engineering knowledge.

That framing puts the system in a supporting role. Patrick Pouyanné, TotalEnergies’ chairman and chief executive, said the goal is tools that help the company’s experts analyze complex data and make decisions. The announcement describes planned models and objectives, not deployed tools or demonstrated operating results.

A joint lab, not an off-the-shelf rollout

The partnership will create a joint scientific laboratory, bringing TotalEnergies’ subsurface teams together with Mistral’s scientific and technological teams. That setup is meant to produce solutions tailored to TotalEnergies’ business needs, rather than simply attach a general AI model to existing data systems.

What the partners are building toward

• Models for exploration, reservoir characterization, and development.
• Scenario generation for potential exploration developments.
• Support for optimizing or extending existing projects.
• Proprietary models inside what TotalEnergies calls a European digital ecosystem.

For Mistral, the deal is a test of whether its models can be adapted to a technically demanding industrial workflow. Arthur Mensch, Mistral’s co-founder and chief executive, said the project demonstrates the company’s ability to support complex industrial processes and customizable solutions that protect intellectual property.

For TotalEnergies, the partnership is also an ownership play. The company says the collaboration strengthens its AI ecosystem so it can create proprietary models while drawing on its strategic subsurface data. Whether the program improves decisions will depend on how well those planned tools work alongside the experts whose knowledge they are meant to use.