OVHcloud has signed a binding agreement to acquire Dragon LLM, a Paris-based developer of specialised generative AI models, and is using the deal to launch an AI lab focused on training and fine-tuning sovereign large language models for European businesses.
Dragon LLM was originally founded as Lingua Custodia to serve the financial sector, building high-performance models designed to run on local infrastructure. The company targets regulated industries where data sovereignty is not optional, and is a winner of the Large AI Grand Challenge organised by the European Commission.
The acquisition is OVHcloud's first and brings a team of fine-tuning specialists into the French cloud provider's ranks alongside the technological building blocks Dragon LLM has developed. The combination is intended to let OVHcloud offer generative AI services for sensitive data, deployable either in the cloud or on-premises, depending on what customers require.
The timing reflects a shift that has been building across European enterprise technology for some time. As AI adoption moves from experimentation into production, businesses in banking, insurance, healthcare, and the public sector are running into a structural problem: the most capable AI infrastructure is concentrated in US hyperscalers, and moving sensitive data to those platforms creates regulatory and political exposure that many cannot accept.
OVHcloud's answer is to build the capability domestically. Dragon LLM describes its work as focused on useful, responsible, and efficient European AI, language that maps directly onto the compliance requirements driving demand from the industries it serves.
The financial terms of the acquisition were not disclosed.
The recap
- OVHcloud signs agreement to acquire Dragon LLM developer.
- Dragon LLM is Paris-based and won EU Large AI Grand Challenge.
- OVHcloud launches an AI lab to train and fine-tune sovereign LLMs.