Daily Brief: Cohere-Aleph Alpha merger creates $20bn transatlantic AI challenger

Today, 24.04.2026

Good morning, Human. The map of European AI just got redrawn overnight, and the cartographers are Canadian.

This merger and the questions it raises about European AI sovereignty are exactly the kind of strategic shifts we'll be unpacking at Human x AI Europe on May 19 in Vienna. If you're navigating these waters, the conversation matters.

The Lead: A Merger That Rewrites the Playbook

For years, the European AI narrative has been one of aspiration and frustration. Build sovereign capacity. Reduce dependence on US hyperscalers. Keep data under European jurisdiction. The problem was always the same: where's the company that can actually deliver at scale?

Today's Cohere-Aleph Alpha merger offers an answer, though not the one Brussels might have scripted. The combined entity will be headquartered in both Canada and Germany, valued at around $20 billion, and explicitly positioned as a "sovereign alternative" to OpenAI, Anthropic, and Google. Both the German and Canadian governments have endorsed the deal.

The mechanism here is worth understanding. Cohere, founded in 2019 and valued at $7 billion in 2025, has built its business on private deployments of large language models (LLMs, the technology underpinning systems like ChatGPT) for enterprise clients. Aleph Alpha, also founded in 2019, was once Germany's flagship LLM developer but pivoted last year from building models to helping businesses and governments use AI. That pivot, which followed the departure of CEO and cofounder Jonas Andrulis, looked like a retreat. Now it looks like positioning.

The financial architecture tells the story. Cohere shareholders will receive approximately 90 percent of the combined company; Aleph Alpha shareholders get around 10 percent. But the strategic value flows both ways. Aleph Alpha brings deep relationships with German industry and government. Cohere brings scale, capital, and a model that European institutions can deploy without sending data to US servers.

Schwarz Group, which owns supermarket chain Lidl and was already a major Aleph Alpha shareholder, is committing $600 million to Cohere's upcoming funding round. This isn't a passive investment. Schwarz has been building data center infrastructure to support AI workloads in Germany. The merger gives that infrastructure a purpose.

The question is whether "sovereign" can compete with "best." OpenAI and Anthropic have the most capable models. Google has the distribution. Cohere's bet is that for a significant segment of the market, control matters more than capability. European governments procuring AI systems, German manufacturers protecting trade secrets, financial institutions navigating GDPR (the General Data Protection Regulation, Europe's data privacy framework): these are customers who will pay a premium for infrastructure they control.

Berlin is expected to act as an anchor customer, prioritizing domestically controlled AI systems in public procurement. If that pattern spreads to other European capitals, the merger creates not just a company but a category.

The Infrastructure Play: Nordic Compute Goes Global

While the Cohere-Aleph Alpha deal grabbed headlines, a quieter story reveals where European AI infrastructure is actually being built. Helsinki-based Verda (formerly DataCrunch) raised $117 million to scale its AI cloud infrastructure, with plans to expand from Finland and Iceland into the UK, US, and Asia.

The numbers are striking. Verda's revenue run rate doubled to over $60 million in Q1 2026. The company is already cash-flow positive. It's one of a select group of NVIDIA Preferred Partners globally, working with clients including Nokia, 1X, ExpressVPN, and Freepik.

The differentiation is clean energy. Verda's data centers in Finland run on 100 percent renewable power, leveraging the Nordics' natural advantages in electricity costs and cooling efficiency. As AI training runs consume increasingly massive amounts of energy, this isn't just a sustainability story. It's a cost story.

CEO Ruben Bryon made a pointed observation to Sifted: "A lot of European companies, at some point, change their headquarters to the US or Delaware. That is not something that we intend to do. We want to stay true to our goals to be a European company and more neutral. Not being a US company is quite important."

The round was led by Lifeline Ventures with participation from byFounders, Tesi (Finnish Industry Investment Ltd, Finland's state-owned venture capital company), and Varma. Debt financing came from Nordic financial institutions. The company plans to hire over 100 people and open offices in California and Asia.

Watch the construction plans. Verda is currently leasing data center space but is building in Finland and expanding to Sweden for larger sites. European AI infrastructure is being built, just not where the policy documents imagined.

The Regulatory Situation: Supervisors Shouldn't Chase Competitiveness

A new CEPS (Centre for European Policy Studies) publication argues against a proposal that's been gaining traction in Brussels: giving financial supervisors a secondary mandate for competitiveness.

The argument is worth understanding because it applies beyond finance. The proposal sounds reasonable: Europe needs a more competitive financial system, so why not task supervisors with promoting competitiveness alongside their primary prudential duties?

CEPS's answer: because it muddies what supervision is meant to do. Competitiveness should be advanced through stronger lawmaking, clearer and more proportionate rules, deeper market integration, and better institutional design. Not by pulling supervisors into political and industry pressures precisely where clarity, independence, and discipline matter most.

The parallel to AI governance is direct. As the AI Act (the EU's comprehensive artificial intelligence regulation) enters enforcement, there will be pressure to make supervisors consider innovation and competitiveness alongside safety and rights. The CEPS argument suggests this is the wrong path. Better to write clearer rules than to ask enforcers to balance competing mandates.

Think Tank Watch: ELLIS Convenes on AI Safety

The ELLIS Workshop on Safe and Secure Artificial Intelligence is scheduled for May 20, 2026. Co-organized by ELLIS units in Lviv, Prague, and Alicante, along with the ELLIS Program on Human-Centric Machine Learning, the virtual workshop will address privacy, safety, trust, biases, and interpretability in AI.

ELLIS (European Laboratory for Learning and Intelligent Systems) is Europe's network of excellence connecting top AI researchers across borders. When ELLIS convenes on safety, it signals where European AI governance thinking is headed. The workshop's focus areas, particularly interpretability and bias, align with the technical requirements emerging from AI Act implementation.

The Lviv unit's involvement is notable. Ukrainian AI researchers remain integrated into European research networks despite the ongoing conflict, a quiet demonstration of how scientific collaboration persists through geopolitical disruption.

The Numbers That Matter

• $20 billion: Valuation of the combined Cohere-Aleph Alpha entity, making it one of the most valuable AI companies globally.
• $600 million: Schwarz Group's commitment to Cohere's funding round, the largest single investment in European AI infrastructure this year.
• 90/10: The ownership split between Cohere and Aleph Alpha shareholders in the combined company.
• $117 million: Verda's new funding round for Nordic AI cloud infrastructure.
• $60 million: Verda's revenue run rate as of Q1 2026, doubled from the previous quarter.
• 100+: New hires planned by Verda, focused on engineering and AI research.

The Week Ahead

• May 20: ELLIS Workshop on Safe and Secure AI (virtual). Registration open for Zoom link.
• Ongoing: Watch for details on the Cohere-Aleph Alpha deal closing and any regulatory review requirements in Germany, Canada, or the EU.
• Infrastructure: Verda's expansion announcements expected as the company opens offices in California and Asia.

The Thought That Lingers

The Cohere-Aleph Alpha merger is being framed as a sovereignty play, and it is. But there's something else happening here. For years, European AI policy has been about defense: protecting data, regulating risks, preventing harms. Today's news suggests a shift toward offense: building alternatives, capturing markets, competing on terms that favor European strengths.

The question is whether "sovereign" becomes a genuine market category or remains a policy aspiration. The answer depends less on what governments say than on what enterprises buy. If Berlin's procurement decisions spread to Paris, Rome, and Warsaw, if Schwarz's infrastructure investment attracts followers, if Verda's clean-energy compute finds customers beyond the Nordics, then something real is being built.

The map is being redrawn. The cartographers may be Canadian, but the territory is European.

Human×AI Daily Brief is compiled from Tech.eu, Sifted, CEPS, and ELLIS. This is meant to be useful, not comprehensive.

Frequently Asked Questions

What is the Cohere-Aleph Alpha merger?

Canadian AI company Cohere is acquiring Germany's Aleph Alpha in a deal that values the combined entity at approximately $20 billion. The merger creates a transatlantic AI company positioned as a "sovereign alternative" to US-based AI providers like OpenAI and Google.

Why is this merger significant for European AI sovereignty?

This is the first major consolidation that explicitly positions "sovereign AI" as a market category rather than just a policy goal. It gives European enterprises and governments a credible alternative to US AI providers while keeping data under European jurisdiction and control.

What role does Schwarz Group play in this deal?

Schwarz Group, which owns Lidl and was already a major Aleph Alpha shareholder, is committing $600 million to Cohere's upcoming funding round. The company has been building data center infrastructure in Germany to support AI workloads, and the merger gives this infrastructure strategic purpose.

How does Verda fit into the European AI infrastructure picture?

Helsinki-based Verda raised $117 million to scale its AI cloud infrastructure, leveraging clean Nordic energy. The company runs data centers on 100% renewable power and is expanding globally while maintaining its European headquarters, representing a different approach to building European AI infrastructure.

What are the implications for AI regulation in Europe?

The merger comes as the EU AI Act enters enforcement. There's growing debate about whether AI supervisors should balance safety and rights concerns with competitiveness goals, with experts arguing that clearer rules are better than asking enforcers to juggle competing mandates.