IA en un minutoNewsBusiness and products
Crédit Agricole teams up with Mistral to bring AI to the whole group
IA en un minuto newsroom · Editor: Jon Elgezabal
In 30 seconds
Crédit Agricole Artificial Intelligence, the bank's AI arm, will have access to Mistral AI's entire catalog, and Mistral will also help put up the group's shared platform. Mistral and the bank are starting with an assistant for customers of its branch networks, automated processes with agents, and cybersecurity. The amount has not been disclosed; in June the group earmarked close to 500 million euros for AI through 2028.
Crédit Agricole Artificial Intelligence, the AI company that the French banking group Crédit Agricole has set up so that all its entities share the same AI tools, has signed a partnership with the French lab Mistral AI. The bank announced it on Tuesday, October 20.
Under the agreement it will be able to use all of Mistral's models, products and tools, and Mistral will also take part in building the AI platform designed to serve all of the group's customers and employees. The first joint projects are a conversational agent (an assistant that talks with customers and helps them with their requests) for customers of the group's retail banks, the transformation of key processes with AI agents, and cybersecurity, which the bank describes as one of the main pillars of the partnership. According to Crédit Agricole, working with Mistral will shorten the delivery times of those projects.
The group also wants to tune large language models to its own data and use cases with Mistral's technology, so that they perform better, consume fewer computing resources and, as a result, have a smaller environmental impact. To distribute AI workloads it will rely on CA-GIP's infrastructure: depending on how sensitive and critical they are, they will go to the group's private cloud, to European partners or to other solutions.
Within the "AI for All" program for its employees, it is also working with the French company Prisme.AI, so that each employee can use, create and share their own AI agents. At the same time it is rolling out Sécur'IA, a platform where teams design and share agents connected to the bank's systems. Its previous tool, Sécuri'Chat, arrived at Crédit Agricole S.A. in 2024 and is now used regularly by nearly 9 out of 10 trained employees. The bank presents these partnerships as a way to support the French and European technology ecosystem and to keep its dependencies under control.
Crédit Agricole has not disclosed the value of the partnership. In June, as reported by ABC Bourse with information from Agefi-Dow Jones, the group announced that it would invest close to 500 million euros in AI between 2026 and 2028, within an IT budget of approximately 5.7 billion euros a year, and that a large share would go to securing computing capacity and infrastructure. At the time it also announced the creation of this AI company, with capital of 150 million euros, due to be up and running in September.
Why it matters · analysis and opinion
European banking is leaving behind the stage of trying out standalone assistants and building its own platform to serve the whole group, and Crédit Agricole is doing it with a European partner, with the most sensitive computing split between its private cloud and nearby providers. It is a way of controlling who it depends on in a sector where customer data cannot go just anywhere. The announcement leaves unclear what will measure success: when the assistant will reach customers and which requests it will actually be able to handle. For other institutions and for those selling them technology, the signal is that a good model is no longer enough: they will have to explain where each task runs and under what control.
Official source: Crédit Agricole · Written with the help of AI: how we make the news
