Billion-dollar deal: Cohere swallows Aleph Alpha – A frontal assault on the US tech giants?

The anti-OpenAI alliance: Why Germany and Canada are now building a joint AI

“Sovereign AI”: This is how Lidl’s parent company and Cohere want to shake up the global AI market

Canadian AI giant Cohere is acquiring the German flagship startup Aleph Alpha. Backed by the governments of Berlin and Ottawa and fueled by massive infrastructure investments from the Schwarz Group (Lidl, Kaufland), a new transatlantic technology alliance is emerging. The ambitious goal: to create a sovereign, highly secure AI ecosystem for businesses and government agencies, offering a genuine, data privacy-compliant alternative to the dominant US hyperscalers like OpenAI and Google. But behind this strategic, multi-billion-dollar deal lies not only a geopolitical declaration of war, but also the sobering story of the failure of the purely European AI dream. Read on to find out why this merger could shape the landscape of artificial intelligence for years to come – and what enormous hurdles this mega-project still has to overcome.

When two nation-states marry a company — and the West finally strikes back

On April 24, 2026, an era came to an end. With the official announcement of the acquisition of the German AI startup Aleph Alpha by the Canadian company Cohere, not only was a company sale completed, but a new direction in global technology policy was established. The transaction, the financial details of which the parties involved are keeping secret, bears the hallmarks of two governments that have recognized that the competition for artificial intelligence is no longer a purely economic phenomenon. Germany and Canada intend to jointly demonstrate that sovereign AI systems are possible beyond the American hyperscalers—and they are relying on a model that combines research, infrastructure, capital, and government support in a single structure.

Two companies, two paths, one goal

The story of both companies begins in the same year, under similar circumstances, at opposite ends of the Atlantic. Jonas Andrulis, an industrial engineer from Heidelberg, a graduate of the Karlsruhe Institute of Technology (KIT), and a former AI manager at Apple's secret lab in Silicon Valley, returned to Germany in 2019 and founded Aleph Alpha together with Deloitte manager Samuel Weinbach. The location was no coincidence: Heidelberg lies within easy reach of the renowned Cyber ​​Valley research network and KIT, both centers of European AI research. The company was intended to be Europe's answer to OpenAI—a European Large Language Model, explainable, transparent, and compliant with data protection regulations, thus the exact opposite of the black boxes from San Francisco.

On the other side of the Atlantic, Google Brain alumni Aidan Gomez, Nick Frosst, and Ivan Zhang founded the company Cohere in Toronto that same year. The name was deliberately chosen: Coherence, as a technical concept of attention mechanisms that Gomez had co-developed as a co-author of the legendary "Attention is All You Need" paper, was to become the company's philosophy—the integration of disparate information into an understandable whole, applied to the needs of businesses. From the outset, Cohere avoided the consumer market and specialized in enterprise solutions: scalable, secure, cloud-agnostic language models for corporations, banks, and government agencies.

Both companies shared a fundamental conviction: the race for AI dominance will not be decided solely by spectacular chatbots, but by trustworthy, controllable systems that meet the demands of regulated markets. This congruence of business models, as it would turn out, was the true strategic foundation for the subsequent merger.

The failure of the European OpenAI dream

To truly understand the merger, one must first trace Aleph Alpha's journey through the highs and lows of the European AI landscape. The company experienced an early surge in popularity: with a funding round announced in November 2023 as a $500 million investment, Aleph Alpha seemed to have finally arrived in the league of global AI champions. Investors such as the Schwarz Group, Robert Bosch Venture Capital, SAP, and several government funding bodies expressed their confidence. But behind the glittering facade lay a critical reality.

Internal documents obtained by Capital magazine painted a sobering picture: The equity stake actually paid into Aleph Alpha's accounts amounted to only around €110 million—the remaining €300 million was earmarked for research and development, to be distributed in phases, and another €60 million consisted of contract commitments. Revenue for 2023 was a mere €945,000. The targeted $20 million in revenue for 2024 was also not achieved, according to the company; estimates had been around €40 million. Losses of nearly €19 million in 2023 reflected the high capital requirements associated with operating and further developing large language models.

Even more significant than the figures themselves was the strategic realignment the company subsequently undertook. Aleph Alpha abandoned direct competition with OpenAI, Google, and Meta in its model development. Instead, the company shifted its focus to so-called sovereign AI solutions: specialized applications for businesses and government agencies that demand the highest standards of data protection, explainability, and regulatory compliance. The Luminous model was increasingly complemented by the PhariaAI product line, an offering for regulated institutions such as hospitals, government ministries, and financial institutions. In January 2025, at the World Economic Forum in Davos, Aleph Alpha unveiled a technologically remarkable innovation: the T-Free architecture, a tokenizer-free language model that significantly improves the efficiency of fine-tuning for less common languages, specialized fields, and diverse writing systems. The collaboration with AMD and the cloud division Schwarz Digits gave this move industrial substance.

But the real drama unfolded beyond the technology. Founder Jonas Andrulis left the company in mid-January 2026. The Schwarz Group acquired Robert Bosch Venture Capital's shares and further increased its stake. The former symbol of German entrepreneurial spirit had lost its independent core and was on its way to becoming an industrial subsidiary of a retail conglomerate—or, as would become clear a few months later, part of an even more ambitious transatlantic construct.

Cohere's Rise: Enterprise AI as a Capital Market Story

While Aleph Alpha struggled to survive, Cohere wrote one of the most impressive growth stories in the enterprise AI segment. The company raised $500 million in an oversubscribed funding round in August 2025, valuing it at $6.8 billion, and followed up with another $100 million in September 2025, bringing its valuation to $7 billion. Investors included AMD, Nvidia, Salesforce Ventures, and the AI-focused venture capital firm Radical Ventures.

The funding rounds reflected a solid business model: Cohere achieved recurring annual revenue of approximately $240 million in 2025, exceeding its self-imposed target of $200 million. The gross margin averaged 70 percent, which is exceptional for a capital-intensive, hardware-dependent AI company and indicates efficient server utilization and strong customer loyalty. Cohere has offices in Toronto, San Francisco, Montreal, London, and Paris and serves clients such as the Royal Bank of Canada, BCE, Oracle, and Dell Technologies. Aidan Gomez had strategically positioned the company as a data-secure alternative to the US hyperscalers and had even traveled to Frankfurt to personally convince European financial institutions of his model.

Cohere's key strategy differed fundamentally from that of major consumer AI providers: The company offers its language models in a cloud-agnostic manner—via public clouds, virtual private clouds, or entirely on-premises—thus giving customers complete control over their data. This architectural decision is non-negotiable for banks, government agencies, healthcare systems, and energy companies, and explains why Cohere is growing in regulated sectors, while open chatbot providers encounter structural resistance there.

The political background: An alliance that was brewing

The merger between Cohere and Aleph Alpha would have been inconceivable without a remarkably consistent geopolitical prelude. In December 2025, Canada and Germany agreed to establish the Canada-Germany Digital Alliance to intensify cooperation in artificial intelligence, digital infrastructure, and quantum computing. This move was anything but symbolic diplomacy: Germany is Canada's largest trading partner within the European Union, and both countries had recognized that their growing reliance on American technology companies posed a strategic risk in light of the geopolitical turmoil of 2025.

On February 14, 2026, on the sidelines of the Munich Security Conference, Canadian Digital Minister Evan Solomon and German Digital Minister Karsten Wildberger signed a joint declaration of intent for AI cooperation and launched the Sovereign Technology Alliance. The message was clear: Canada and Germany no longer wanted to act as passive consumers of American platforms in the field of AI, but rather to establish their own independent technological sovereignty. Wildberger stated it openly: They needed their own path, different from that of the US, and this path led through partnerships. For Canada, the trade policies of the Trump administration also played a role: The threat of 100 percent tariffs on Canadian goods and the growing pressure from US protectionism drove Ottawa into the arms of European partners.

Given these circumstances, the merger of Cohere and Aleph Alpha was no longer a surprise, but rather the logical industrial policy consequence of a diplomatically prepared process. What began in February as a declaration of intent at the ministerial level landed on the desks of corporate lawyers in April.

Anatomy of the deal: Who gets what?

On April 24, 2026, both companies officially announced the merger. The formal press conference in Berlin was attended by Digital Minister Wildberger and his Canadian counterpart, Evan Solomon—an unusual ritual in the company's history, underscoring the governmental nature of the transaction. The merged company will be called Cohere, and Aidan Gomez will remain CEO. It is an acquisition marketed as a merger.

The ownership structure is clear: Cohere shareholders will receive approximately 90 percent of the shares in the merged company, while Aleph Alpha shareholders will receive about 10 percent. The financial valuation of the entire structure is stated at around US$20 billion, a figure that includes the planned financing round and is still subject to market confirmation. The purchase price for Aleph Alpha itself has not been disclosed, which is hardly surprising given that the company was recently valued at less than half a billion euros and still has relatively small revenues.

The decisive element of the deal is the Schwarz Group's involvement. The retail giant from Neckarsulm, which owns Lidl and Kaufland and was already an investor in Aleph Alpha, will contribute $600 million to Cohere's upcoming funding round. This capital is not just a financial transaction, but a strategic alliance: The merged company will use the Schwarz Group's data centers for its AI services. This gives Cohere access to one of the most ambitious European cloud infrastructures—and secures the Schwarz Group the role of preferred infrastructure partner for the new global sovereign AI champion.

The infrastructure bet: Schwarz Digits as a foundation

The Schwarz Group, known in Germany primarily as the operator of Lidl and Kaufland, has developed into a serious technology player in recent years. In November 2025, the company laid the foundation stone for a new data center in Lübbenau in the Spreewald region—with an investment volume of eleven billion euros, the largest single investment in the company's history. The Schwarz Digits data center will be equipped with up to 100,000 AI-specific GPUs, thus achieving a computing capacity that far exceeds that of the planned Telekom-Nvidia data center in Munich with its 10,000 GPUs. The facility will be powered by renewable energy, and its waste heat will be fed into the region's district heating network—a model that combines economic and environmental goals.

This infrastructure is the foundation upon which the new Cohere-Aleph-Alpha company can deliver on its promises to European customers. The promise is this: their AI stack remains in Europe, is subject to European law, and no US corporation or Chinese state has access to it. For government ministries, hospitals, energy companies, and financial institutions in regulated markets, this guarantee is not just a selling point, but often a legal necessity. The General Data Protection Regulation (GDPR), European AI legislation, and sector-specific regulations like DORA in the financial sector enforce precisely the kind of data localization that US hyperscalers can only offer with difficulty, or not at all, due to structural limitations.

For the Schwarz Group, this model is also strategically consistent. Whoever controls data centers in Europe controls a key bottleneck in the AI ​​economy: computing infrastructure. By simultaneously being the main investor, infrastructure provider, and anchor customer of the new company, the Schwarz Group secures a position in the AI ​​ecosystem that is virtually unassailable—and it does so through capital rather than research.

A new dimension of digital transformation with 'Managed AI' (Artificial Intelligence) – Platform & B2B solution | Xpert Consulting - Image: Xpert.Digital

Here you will learn how your company can implement customized AI solutions quickly, securely and without high entry barriers.

A managed AI platform is your all-inclusive, worry-free solution for artificial intelligence. Instead of dealing with complex technology, expensive infrastructure, and lengthy development processes, you receive a ready-made solution tailored to your needs from a specialized partner – often within just a few days.

The key advantages at a glance:

⚡ Rapid implementation: From idea to ready-to-use application in days, not months. We deliver practical solutions that create immediate added value.

🔒 Maximum data security: Your sensitive data stays with you. We guarantee secure and compliant processing without sharing data with third parties.

💸 No financial risk: You only pay for results. High upfront investments in hardware, software, or personnel are completely eliminated.

🎯 Focus on your core business: Concentrate on what you do best. We take care of the entire technical implementation, operation, and maintenance of your AI solution.

📈 Future-proof & scalable: Your AI grows with you. We ensure continuous optimization and scalability, and flexibly adapt the models to new requirements.

More information here:

• The Managed AI Solution - Industrial AI Services: The Key to Competitiveness in the Services, Industry and Mechanical Engineering Sectors

Strategic logic: What each side gains

The deal follows a clear economic logic that is convincing for both sides when one honestly considers the initial conditions.

Cohere is primarily gaining market access and credibility in Europe. The company recognized that Europe and the United Kingdom are its fastest-growing markets, as Gomez personally stated to Handelsblatt. However, without established relationships with European regulators, in-depth knowledge of local compliance requirements, and a presence on European infrastructure, even the best enterprise AI product reaches its structural limits. Aleph Alpha provides all of this: years of experience with the German Federal Employment Agency, Deutsche Bank, Bosch, and several state administrations. Furthermore, it positions itself as a company that explicitly embraces European values ​​such as transparency, regulatory compliance, and explainability as part of its product promise.

Aleph Alpha gains primarily in scalability. The Heidelberg-based company's fundamental problem wasn't the technology itself, but rather its ability to commercialize it at a global pace. Enterprise AI sales cycles are long, sales structures for multinational corporations are capital-intensive, and obtaining international certifications ties up management resources. Cohere brings not only $240 million in annual recurring revenue, but also the operational experience of serving large enterprise customers across multiple continents simultaneously. The combined platform is intended to serve customers in energy, defense, finance, telecommunications, healthcare, and the public sector—a market spectrum that Aleph Alpha alone could never have covered.

Both companies enjoy a level of political backing rarely found in the private-sector AI industry. When two nation-states act as anchor customers, partners, and diplomatic advocates, it significantly reduces customer acquisition costs. Governments that publicly endorse a provider signal to other governmental and semi-governmental institutions that this provider minimizes the political and regulatory risk of a procurement decision.

The market environment: Europe's structural weakness and the alternative

To understand the significance of the deal, one must consider the initial situation it is competing against. A KPMG study from January 2026, conducted jointly with the German AI Association and based on over 900 expert interviews, found that the USA clearly leads the global AI race in all dimensions examined. While Europe follows ahead of China, it cannot keep pace with the economic dynamism of the USA. US providers dominate the European cloud market; Amazon Web Services, Microsoft Azure, and Google Cloud together account for the majority of data processed in Europe. In terms of AI computing capacity, the USA has many times the capacity of Europe and controls the lion's share of the world's high-end GPU resources.

Added to this is structural fragmentation: Europe lacks a unified digital single market that allows technology companies to scale quickly. Language diversity, differing national data protection laws, and separate procurement processes significantly increase operating costs for European AI providers. Aleph Alpha's T-Free architecture was precisely the answer to this—a model that can handle underrepresented languages ​​more efficiently, thus turning Europe's linguistic fragmentation into an advantage rather than a hindrance. But even technological innovation doesn't automatically succeed in the market without the necessary commercialization structures.

The merger of Cohere and Aleph Alpha is more than just a deal: it's a counter-model. The premise is that companies in regulated markets—which ultimately includes all strategically important sectors of a modern economy—are willing to pay a premium for control, transparency, and sovereignty. Whether this premise holds true will become clear in the quarterly reports of the next two to three years.

The state as entrepreneur: Opportunities and risks of the anchor customer logic

A key element of the deal is the role of the German government as an anchor customer. Federal Digital Minister Wildberger publicly described the project as a "strong signal for Germany as an AI hub" and as the birth of a "German-Canadian AI model: secure, sovereign, and competitive." Berlin is also considering a direct financial stake in the merged company, which would elevate government involvement to a new level. This constellation has historical precedents—one need only think of Airbus as a European defense project—and is not unusual in the current technology policy debate surrounding strategic autonomy.

Nevertheless, close ties to the state entail economic risks that cannot be ignored. When the state acts as the primary customer, a structural dependency arises that can distort incentives for innovation. A company that generates its revenue primarily from public contracts is subject to different incentives than one that must compete fiercely for private enterprise clients. Bureaucratic procurement cycles slow down the feedback loop between the market and the product. And political backing can disappear as quickly as it appeared with a change of government. The new company's promise to build a broad enterprise customer base that extends beyond government clients is therefore not merely a commercial goal, but a strategic prerequisite for survival.

Added to this is the question of regulatory hurdles. The deal requires official approval, and it is expected that both EU competition authorities and the relevant bodies in Canada and possibly the US will examine whether the transaction creates dominant market structures or constitutes legitimate state intervention in strategic infrastructure. The answer to this is politically open and legally complex.

The technological promise: Sovereign AI between aspiration and reality

Behind the buzzword "Sovereign AI" lie concrete technological requirements that the merged company must meet to maintain credibility with its customers. In the context of AI, sovereignty means: the training data remains under the client's control, the models are operated on infrastructure subject to European law, updates and security patches are delivered without access by third countries, and the system's decisions are traceable and auditable. Handelsblatt has precisely defined what this means: control over the entire value chain, from the training data to the models and the cloud infrastructure.

In its press conference, Cohere explicitly pledged to utilize European infrastructure and comply with applicable sovereignty requirements. The foundation for this is the infrastructure of Schwarz Digits. Cohere's CFO, Francois Chadwick, underscored this commitment to Reuters unequivocally: "We will commit to using European infrastructure and complying with the sovereignty requirements applicable here." However, promises and implementation often diverge in the technology sector. The real test will be the ability to deliver technical excellence on a global scale while simultaneously adhering to the operational constraints of a tightly regulated European deployment environment.

Aleph Alpha's T-Free architecture is not a marginal technological detail, but a potential first-order competitive advantage. If language models can be adapted more efficiently to new languages, writing systems, and specialized domains without tokenizers, this significantly reduces fine-tuning costs—and thus the barrier to entry for medium-sized businesses and smaller government institutions that previously could not afford highly specialized models. This could be a crucial lever for market penetration in Europe, where small and medium-sized enterprises (SMEs) play an economic role unparalleled in the US.

Risk analysis: What could jeopardize the deal

No deal of this magnitude is without structural risks. For the merged company, four challenges are particularly serious.

First, there's the valuation issue. With a projected total valuation of around $20 billion after the completion of the new funding round, the company is in a league that entails intense scaling expectations. Cohere's current ARR of $240 million implies a price-to-revenue multiple of around 83x—a high figure that will need to be defended with robust growth and increasing margins. A tech winter, like the one many startups experienced in 2022/23, could put this valuation under significant pressure.

Second, the integration and cultural challenges remain significant. Aleph Alpha is a European research company with a corporate culture focused on scientific rigor, regulatory compliance, and explainability. Cohere is a North American scale-up with a commercial go-to-market focus and rapid iteration cycles. Synthesizing these cultures is no trivial management task, and integration failures are more common than successes in the technology industry.

Third, there is the reputational risk. Aleph Alpha has faced criticism multiple times throughout its history, beginning with the transparency issues surrounding its 2023 funding round. The public perception of a company once marketed as Germany's OpenAI, now being acquired by a Canadian competitor, can be interpreted as a failure—even though the reality is considerably more nuanced. The political narrative that both governments have constructed around the deal also serves to reframe this perception.

Fourth, the company faces overwhelming and relentless competition. OpenAI, Google DeepMind, Anthropic, and Microsoft are investing billions in precisely those enterprise capabilities that Cohere and Aleph Alpha claim as their unique selling proposition. And with Mistral from France, also a European champion and receiving government funding, the new company shares the sovereign AI market in Europe with a well-capitalized competitor.

The big picture: AI as a geopolitical infrastructure problem

The merger of Cohere and Aleph Alpha cannot be adequately described without placing it within the context of a profound shift in the geopolitical coordinate system. Artificial intelligence has become a fundamental building block of economic competitiveness and national sovereignty, much like railways, energy grids, or the internet once were. Whoever controls the AI ​​infrastructure will, in the long run, control access to the most productive sectors of the modern economy.

In this context, the US, with its hyperscalers, occupies a position that is historically unparalleled: it provides the dominant platform on which a large portion of global data processing takes place, and it does so under US law—with all the access rights granted to the US government by the Cloud Act and similar legislation. For European democracies that insist on the rule of law and data sovereignty, this is a structural problem that cannot be fully solved by even the best General Data Protection Regulation (GDPR) as long as the physical infrastructure remains in foreign hands.

The German-Canadian collaboration is the most concrete attempt to date at addressing this dilemma, one that goes beyond government regulation and relies on commercial market power. The underlying premise is compelling: a globally scalable company offering sovereign AI services on its own European infrastructure, covering the full spectrum from enterprise applications to government platforms, creates structural alternatives that no regulatory framework alone can generate. Whether this construct is strong enough to compete with the financial might of American Big Tech remains to be seen. But it is the most serious model that the non-American West has put forward so far.

Canada's Digital Minister Solomon summed it up perfectly: This is just the beginning of a broader initiative for sovereign AI. Wildberger spoke of documenting a different path—one beyond the US, through partnerships. In a world reorganizing itself after an era of digital unilateralism, this is more than rhetoric. It's an experiment in industrial sovereignty—and its outcome will shape the map of the global AI economy for a decade.

Konrad Wolfenstein

I would be happy to serve as your personal advisor.

contact me at wolfenstein ∂ xpert.digital

Just call me on +49 7348 4088 965 .

LinkedIn
 

The quasi-in-house solution: How Xpert.Digital closes operational gaps in B2B marketing and sales – Smart Content-Driven Business - Image: Xpert.Digital

Xpert.Digital is a data-driven B2B industry hub led by Konrad Wolfenstein . The company acts as an external, quasi-in-house solution for industrial partners, closing operational gaps in marketing, content, and sales – without requiring additional resources on the client side.

More information here:

• The quasi-in-house solution: How Xpert.Digital closes operational gaps in B2B marketing and sales – Smart Content-Driven Business