Sesterce, a French company that builds and operates large scale computing capacity for artificial intelligence, announced a major investment plan on October 8, 2026 to develop an AI data centre campus at the former Kaipola paper mill in Jamsa, central Finland. The company said the overall programme could exceed 10 billion euros and would begin with a first phase targeting roughly 200 megawatts of power capacity. Sesterce described the project as the start of a long term effort that could expand to over 600 megawatts in Finland and to larger gigawatt scale ambitions in the years ahead.
What the announcement covers
The plan unveiled by Sesterce sets out a staged approach. The first phase would convert the decommissioned industrial site and bring an initial 200 megawatt facility online, with construction works slated to begin in 2026 according to the company statement. Sesterce emphasised modular design elements and adaptive reuse of the industrial campus, alongside plans to scale up in line with market demand.
In parallel, Sesterce reiterated earlier commitments to deploy AI infrastructure inside France. The company has already proposed and in some cases started smaller projects at several French sites, including a previously announced Valence project and a joint venture in Hauts de France with major partners. Sesterce said its long term aim is to deploy more than one gigawatt of AI compute capacity in Europe over several years, a target that would place it among the largest independent builders of AI infrastructure on the continent if realised.
Why Sesterce chose Finland
Finland has become a magnet for large scale data centre projects because of abundant grid capacity, cool climate that reduces cooling costs, and the availability of industrial sites suitable for rapid conversion. Sesterce framed the Jamsa campus as a fit with the region’s industrial heritage and as an opportunity to deliver construction jobs in the near term, while positioning the site for high density GPU clusters needed by modern generative AI workloads.
For French AI infrastructure builders, locating large data halls in Nordic countries is a common pattern that balances the energy and grid constraints they face at home with the urgency to deliver capacity. Sesterce’s own disclosures show the company is pursuing a two track strategy, building local French capacity while placing very high density deployments where electricity and cooling conditions are most favourable.
Investment scale, partners and next steps
Sesterce described the more than 10 billion euro figure as a programme scale rather than a single contract value, and said financing milestones, site purchases, and permitting remain subject to negotiation and regulatory approval. The company has previously announced partnerships and staged investments in France, and it confirmed discussions with potential investors and industry partners for the Finnish campus.
Officials in Jamsa and regional development agencies will need to review the sale of the former paper mill site and issue the necessary permits. Sesterce said project timelines will depend on those local processes and on securing long term power agreements. The company projects the first tranche of construction would start in late 2026 with incremental expansion thereafter.
How this matters to France and to European AI supply chains
Sesterce’s announcement speaks to three linked trends. First, demand for concentrated GPU compute is driving new industrial scale facilities that require hundreds of megawatts of power. Second, European and French ambitions to host sovereign or locally controlled AI capacity are colliding with practical limits on where that capacity can be placed quickly and sustainably. Finally, commercial players are responding by combining domestic projects with Nordic and other international sites that provide the scale and energy profiles required by present generation models.
For France, Sesterce’s dual strategy underscores the tension in national industrial policy. French officials have emphasised the need for domestic AI factories and sovereign tooling for both public and private use. At the same time, firms face permitting, grid and environmental constraints at home that slow the pace of multi hundred megawatt builds. The Finnish campus signals that French companies are prepared to pursue a hybrid footprint to meet near term market demand while continuing to invest in French projects.
Environmental and community questions
Large data centre projects frequently raise questions about local environmental impact, water use, and the carbon intensity of electricity. Sesterce’s statements highlight reuse of an industrial site as an environmental advantage, and the company has said it will pursue energy sourcing solutions consistent with local regulations and decarbonisation goals. Nevertheless, past French data centre proposals by the same company and others have encountered regulatory scrutiny and legal challenges related to environmental assessment and community opposition, indicating those issues could reappear as Sesterce advances the plan.
Where this leaves the market
If realised at scale, Sesterce’s Jamsa campus would join a growing cluster of hyperscale and specialised AI facilities in the Nordic region and in parts of Europe. That would expand options for European AI developers seeking lower latency alternatives to US hyperscalers or captive national systems. For French AI and infrastructure strategy, the development highlights the need to accelerate permitting, grid planning and industrial coordination if Paris wants more of the high density compute footprint to remain inside France.
Sesterce said it will continue public engagement with Finnish authorities and local stakeholders as the project moves through permitting, and it will make further financing announcements in the months ahead as partners are finalised.
