Naval Group invests in the Thales cortAIx initiative

France’s Naval Group has invested in cortAIx France, a Thales initiative to advance the development of artificial intelligence for defence. The two defence giants will be collaborating on the development of AI for naval applications including command and control and electronic warfare. 

By Sam Cranny-Evans, published on February 12, 2026. 

The Naval Group has invested capital in cortAIx, a company established by Thales to accelerate the development of artificial intelligence (AI) for defence, according to a February 10 press release. 

As part of the deal, the Naval Group has acquired a 20% stake in the company, which has offices in France, the UK, Germany, Canada and Singapore. It will also join the governance board of cortAIx, and a dedicated centre for AI research is to be opened near its facility in southern France. </p>

<span style=”font-weight: 400;”>The goal of both companies is to accelerate the adoption and industrialisation of AI developed within cortAIx. “This unprecedented alliance will accelerate the deployment of AI technologies and their adoption by the armed forces, while upholding algorithm sovereignty and safeguarding sensitive data.”

Specifically, the two companies will work on AI applications for:

• Collaborative combat
• Electronic warfare
• Decision support 
• Training and simulation
• Logistics and maintenance. 

The sensitive data mentioned above is a hot topic, and few areas are more closely guarded than maritime data. Most navies will have extensive reserves of data on the acoustic signatures of threats relevant to them. That data is essential for training AI algorithms that can process sonar data and produce meaningful outputs. But, it is usually very closely guarded and held at the highest levels of classification. Accessing that data can be the difference between success and failure in maritime AI. 

Calibre comment: Why do we need AI at sea?

AI is essential for the jump from automated to autonomous. An automated system is usually hard-coded to respond to a given input in a given way. It can be thought of as a flow chart with inputs and outputs. A good, albeit basic example, is a thermostat. If the temperature falls below a certain number, turn the heating on. When it reaches that number, turn the heating off. 

This type of automation is already used in defence to filter targets from radar screening, for example. And in some cases like the Phalanx close-in weapon system, the weapon can technically engage a target autonomously. The Aegis air defence systems used by US Arleigh Burkes can be fully automated, conducting the air defence of a ship or ships based on its own assessment of the threat. The challenge here is context. If the Aegis system misinterprets the threat based on its extensive but ultimately limited coding, it might empty the ship’s magazine on the first engagement. Not good when reloading requires a trip back to port. 

In theory, AI will be able to make informed decisions that are closer to those a human would make. It is also able to process sensor data in a more complex way, allowing for shades of grey. With AI, defensive systems would no longer be using the “if this, then that” logic, but an array of options based upon specialised training specific to that task. This development is key to autonomy, as well as advanced electronic warfare and C2 in all domains, but a difficult key to cut. Which is likely why Thales and the Naval Group are cooperating and investing in cortAIx. 

If you would like to read more about some of the software powering autonomy, check the following links. It really helps us out, thank you: 

• Applied Intuition UK and Kraken partner on autonomous vessels – Calibre Defence 
• Arondite raises $12 million in seed funding – Calibre Defence 
• Hivemind: ST Engineering and Shield AI partner on autonomous flight – Calibre Defence

The lead image shows Pierre Éric Pommellet, Chairman and CEO of Naval Group, and Patrice Caine, Chairman and CEO of Thales. Credit: Thales/Anthony Guerra.