Today, 24.03.2026
Good morning, Human. The week opens with a quiet structural shift in how Europe funds its AI ambitions. While the headlines chase mega-rounds and foundation model valuations, something more interesting is happening in the plumbing: a single investor in London just closed the largest solo general partner (GP) venture fund ever raised in Europe, and the thesis behind it says as much about where European AI is heading as any billion-dollar raise.
The Lead: Air Street's $232 Million Bet on Concentrated Conviction
Nathan Benaich's Air Street Capital announced the close of Fund III at $232 million on Monday, a figure that makes it the largest solo GP venture fund ever raised in Europe. The milestone is not merely a record for one investor; it represents a structural argument about how AI-focused capital should be deployed.
The fund will write initial cheques of $500,000 to $15 million for early-stage AI-first companies in North America and Europe, with selective growth-stage allocations reaching up to $25 million. Air Street now has $400 million in assets under management, according to The Next Web, up from the $17 million raised for Fund I back in 2020.
Here's the mechanism hiding under the headline: for most of the past decade, the received wisdom in European venture capital was that serious funds required partnership structures – large teams, committee decision-making, distributed accountability. Benaich has spent five years dismantling that assumption. The solo GP model enables faster term sheet execution, consistent investment philosophy across fund cycles, and the ability to pursue contrarian bets that larger partnerships might decline. The tradeoff is concentration risk on the human side – there is no committee to catch a blind spot.
The portfolio tells the story of what AI-first means in practice. Synthesia, the AI video platform, now generates more than $150 million in annual recurring revenue and counts customers across more than 90% of the Fortune 100. Black Forest Labs, whose FLUX models have achieved widespread adoption among developers building visual applications, sits alongside Poolside, a frontier AI lab serving enterprise and government clients. Defence appears explicitly in the fund's mandate – Air Street's portfolio includes Delian Alliance Industries, signaling willingness to operate in sectors where capital requirements, procurement timelines, and regulatory constraints make most generalist funds nervous.
The deeper question is whether Europe can produce the volume of AI-first companies required to absorb the growing supply of capital. The continent has made genuine progress in foundation model research, applied AI for enterprise, and defence technology, but the pipeline of companies capable of scaling globally at speed remains thinner than in North America. Benaich's bet, implicitly, is that it will thicken fast enough.
The Funding Picture: Interloom and the Tacit Knowledge Problem
Munich-based Interloom raised $16.5 million (€14.2 million) in a seed round led by DN Capital, with participation from Bek Ventures and existing investor Air Street Capital. The company is building what it calls a context graph – a continuously updated map of how operational decisions actually get made inside an enterprise, constructed by ingesting millions of real cases: support emails, service tickets, call transcripts, work orders.
The thesis addresses a genuine bottleneck in enterprise AI deployment. Founder and CEO Fabian Jakobi argues that about 70% of operational decisions have never been formally documented. When a complex support ticket lands on a veteran staffer's desk, they know the workaround, the right internal team to escalate to, and the resolution – not because it's in a manual, but because they've seen it before. AI agents, no matter how capable, are useless in large enterprises without this organization-specific context.
The investor lineup carries its own thesis-confirming logic. Guy Ward Thomas, the DN Capital partner leading the investment, was previously the first institutional backer of Cognigy, the German enterprise conversational AI platform acquired by NICE in August 2025 for $955 million – described at the time as Europe's largest AI exit. Mehmet Atici of Bek Ventures previously backed UiPath, the robotic process automation pioneer. Both have learned that context is what makes AI agents work in practice.
Interloom's early customer base includes Zurich Insurance, JLL, Fiege, Commerzbank, and Volkswagen. At Commerzbank, the company reportedly reduced the gap between documented and actual operational knowledge from roughly 50% to 5%. The timing may be propitious: roughly 10,000 Baby Boomers retire daily in the U.S., walking out the door with decades of institutional knowledge just as companies are trying to deploy AI at scale.
The Infrastructure Play: Inside Paris's Underground Quantum Lab
Tucked in the Latin Quarter of Paris, near the Panthéon, quantum startup C12 has built a 1,000 square metre underground laboratory where it assembles quantum computers using carbon nanotubes – a material the company believes will overcome the limitations of silicon-based approaches.
C12's approach uses isotopically purified carbon-12 nanotubes to host spin qubits with exceptionally low noise. By suspending these nanotubes above silicon chips, the company maximally isolates qubits from environmental perturbations, achieving coherence times two orders of magnitude larger than previously measured in carbon-based quantum circuits. The company recently partnered with QC Design to adopt Plaquette software for simulating and optimizing fault-tolerant quantum computing designs.
The Paris location is deliberate. France has positioned itself as a European quantum hub, with 80% of French quantum research conducted between Paris, Saclay, and Grenoble. The region hosts Pasqal, Quandela, Alice & Bob, and now C12 – each pursuing different quantum computing architectures. C12 was spun out of the Physics Laboratory of the École Normale Supérieure in 2020 by twin brothers Pierre and Matthieu Desjardins, and has raised over €28 million in equity funding while building a diversified portfolio of more than ten patent families.
The underground facility matters because quantum computers require extreme isolation from noise, light, and vibration. Building in the heart of a major city while achieving the necessary environmental control is an engineering statement as much as a scientific one.
The Numbers That Matter
$232 million – Air Street Capital's Fund III, the largest solo GP venture fund ever raised in Europe (TechCrunch)
$16.5 million – Interloom's seed round for AI agent knowledge infrastructure, led by DN Capital (Fortune)
70% – Estimated share of operational decisions in enterprises that have never been formally documented, per Interloom's analysis (The Next Web)
$21.8 billion – European AI funding in 2025, up 58% year-over-year, according to a new Prosus and Dealroom.co report
133 million – Monthly LLM users in Europe, nearly double the US figure, though almost every model they use was built in America or China (Prosus)
73% – Share of lead investors in late-stage European AI companies that are American (Prosus)
1,000 sq m – Size of C12's underground quantum laboratory in central Paris (Sifted)
Think Tank Watch: Europe's Invisible Giant
A new report from Prosus and Dealroom.co titled "State of AI in Europe: The Invisible Giant" surfaces three structural contradictions that define Europe's position in the AI race.
The Talent Paradox: Europe has roughly 325,000 AI professionals – the same number as the United States. Three of the ten most-cited AI scholars globally are European. Yet 53% of European AI talent works in traditional economy roles such as industrial groups and consultancies, compared to 40% in the US, and six of Europe's top fifteen AI employers are US Big Tech companies.
The Usage Paradox: Europeans are the world's most active AI users. Europe's 133 million monthly LLM users nearly doubles the US figure. But almost every model they use was built in America or China – meaning European users are, in effect, training and improving algorithms owned by others.
The Startup Paradox: Europe creates as many new AI startups each year as the US. But it converts them into breakout companies at a third of the rate. By the time a European AI company reaches late stage, 73% of its lead investors are American. Europe incubates; the US captures.
The report argues this is not a failure of talent or innovation – but a failure of capital and structural ambition. And crucially, it argues both are fixable.
The Week Ahead
March 31 – Quantum Night in Paris at the Cité des Sciences, La Villette, hosted by the French Physics Society in partnership with the Cité des Sciences
Ongoing – 28DIGITAL's €60,000 open call for incubators to support Europe's next generation of AI startups remains open
Watch – Station F's new F/ai accelerator, backed by Sequoia, General Catalyst, Mistral and OpenAI, has announced its first batch of 20 AI companies
The Thought That Lingers
There's something almost paradoxical about Air Street's milestone. The largest solo GP fund in European history was raised to back AI-first companies – a category defined by the ability of machines to learn, reason, and act autonomously. Yet the fund's entire thesis rests on the irreducible value of a single human's judgment: one investor's ability to see patterns, take contrarian positions, and move faster than committees.
The same tension runs through Interloom's pitch. The company exists because AI agents cannot function without the tacit knowledge that lives in human experts – the veteran staffer who knows why the documentation is wrong, the underwriter who understands their broker's risk appetite better than any consultant ever could. The most sophisticated AI systems still need to learn from the most human forms of expertise.
Perhaps that's the real signal in today's news. The AI ecosystem is maturing not by replacing human judgment, but by finding new ways to capture, preserve, and scale it. The question is whether Europe can build the capital structures and institutional memory to do the same for its own innovation ecosystem – before someone else captures the value.
These questions are already on the agenda. Human x AI Europe, May 19, Vienna, is where Europe's AI ecosystem sits down and gets serious about the answers.
Human×AI Daily Brief is compiled from TechCrunch, Fortune, The Next Web, Sifted, Prosus, The Quantum Insider, and EU-Startups. This is meant to be useful, not comprehensive.
Frequently Asked Questions
Q: What is a solo GP venture fund?
A: A solo GP (general partner) venture fund is managed by a single decision-maker rather than a partnership of multiple investors. Air Street Capital's $232 million Fund III is now the largest such fund ever raised in Europe, enabling faster investment decisions and consistent thesis execution.
Q: What is tacit knowledge in enterprise AI?
A: Tacit knowledge refers to expertise that has never been formally documented – the institutional memory held by experienced employees who know workarounds, escalation paths, and solutions from experience rather than manuals. Interloom estimates 70% of operational decisions fall into this category.
Q: How does C12's quantum computing approach differ from competitors?
A: C12 uses isotopically purified carbon-12 nanotubes suspended above silicon chips to create spin qubits. This approach achieves coherence times two orders of magnitude larger than previously measured in carbon-based quantum circuits, according to the company's partnership announcement with QC Design.
Q: What is Europe's AI funding gap with the United States?
A: According to the Prosus and Dealroom.co report, Europe creates as many new AI startups annually as the US but converts them into breakout companies at one-third the rate. By late stage, 73% of lead investors in European AI companies are American.
Q: What is a context graph in enterprise AI?
A: A context graph, as developed by Interloom, is a continuously updated map of how operational decisions actually get made inside an organization. It's constructed by ingesting millions of real cases – support emails, service tickets, call transcripts – to capture how expert workers actually solve problems.
Q: When is the EU AI Act high-risk compliance deadline?
A: The August 2026 deadline for high-risk AI system compliance under the EU AI Act remains the key regulatory milestone. Companies deploying AI in areas like employment, credit scoring, and critical infrastructure must demonstrate conformity by this date.