Europe’s sovereign-AI market is moving from a collection of national champions and infrastructure projects towards consolidation, with Cohere and Germany’s Aleph Alpha signing the definitive agreement for a combination first outlined earlier this year.

The enlarged company will operate globally under the Cohere name, with dual headquarters in Berlin and Toronto and more than 1,000 employees across Europe and North America. Aleph Alpha’s Heidelberg operation is due to remain a research centre, while completion remains subject to regulatory approvals and is expected later in 2026.

The agreement gives Cohere a larger institutional base in Germany while bringing one of Europe’s better-known domestic model developers into a company with wider international distribution. Aleph Alpha, meanwhile, gains access to Cohere’s commercial reach at a point when sovereignty is becoming part of procurement criteria for governments and regulated industries rather than remaining principally a policy slogan.

Leadership will also change after completion. Aleph Alpha co-chief executive Ilhan Scheer is set to become Cohere’s chief operating officer, while co-founder and co-chief research officer Samuel Weinbach will take the chief research officer role at the combined company.

Sovereignty becomes an infrastructure question

The companies are presenting the combination around control of data, technology, and deployment rather than solely around model performance. Cohere already sells systems designed to run across public cloud, private infrastructure, on-premises environments, and air-gapped deployments, while Aleph Alpha has concentrated much of its commercial strategy on government and regulated customers in Europe.

Aidan Gomez, Cohere’s co-founder and chief executive, said: “No government or enterprise should have to choose between capable AI and control over their technology.” Delivering that control, however, depends on considerably more than the headquarters address of the model provider.

Data residency, cloud ownership, operational jurisdiction, security controls, model access, and the ability to move workloads between suppliers all feed into sovereignty claims. An organisation can buy a European-labelled model while remaining dependent on foreign compute, proprietary infrastructure, or externally controlled software layers.

The combined business intends to deepen its work with STACKIT, the sovereign cloud platform operated by Schwarz Digits, the technology arm of the group behind Lidl and Kaufland. That gives the company a route to pair model technology with European cloud infrastructure for customers seeking to reduce reliance on the largest US hyperscalers.

The resulting proposition sits between buying frontier AI from a global model company and procuring a nationally contained stack. Cohere remains a Canadian-founded international business, but the planned structure gives Germany a substantial operational and research role rather than treating Europe principally as a sales territory.

Consolidation tests Europe’s AI economics

The combination also reflects the economics facing independent model developers. Training and serving increasingly capable systems demands large amounts of computing infrastructure and capital, while enterprise customers expect security, integration, support, compliance, and deployment flexibility on top of the underlying model.

Aleph Alpha has built political visibility and public-sector relationships in Germany, although sovereign credentials alone do not remove the need for commercial scale. Cohere brings a larger international customer base and more capital, while Aleph Alpha contributes research capacity, German institutional relationships, and a stronger position inside Europe’s digital-sovereignty debate.

The deal does not remove the tension between technological independence and commercial concentration. If Europe’s answer to dependence on a handful of American AI platforms is to consolidate smaller providers into fewer larger ones, customers may gain stronger alternatives while still facing a market where meaningful choice requires considerable scale.

Competition will also extend beyond model developers. European policy is directing public and private investment towards AI factories, supercomputing, cloud capacity, chips, and data infrastructure, while companies including Schwarz Digits are trying to turn those assets into commercially viable platforms. A sovereign-AI supplier therefore has to fit into a broader technology chain rather than compete solely on model benchmarks.

Cohere and Aleph Alpha are betting that a transatlantic structure can reconcile those pressures by combining Canadian AI development with a substantial German base and European infrastructure partnerships. Procurement teams will ultimately test that proposition through practical questions: where data sits, which legal regimes apply, who operates the systems, and whether customers retain control when requirements change.

Regulatory approval and integration still stand between the agreement and the finished company, and the businesses have not detailed how their product portfolios will be combined. Even so, the definitive deal takes sovereign AI another step away from a fragmented market of national contenders and towards a contest over which providers can assemble enough models, infrastructure, capital, and institutional trust to operate internationally.