Week 8
From Paris to New Delhi: Macron's Two Faces of AI, Europe's $3B Startup Surge, and Yann LeCun's Contrarian Bet Against LLMs.
An incredibly dense week, driven by the AI Impact Summit in New Delhi—the first global AI summit hosted by a Global South nation—and a flurry of European moves that prove the continent is done talking and has started building.
Let’s start with the most striking observation: Emmanuel Macron’s two speeches, two strategies.
At the Paris AI Action Summit (February 2025), Macron played the business card: €109 billion in investment pledges, a call to make France the “world’s No. 3 in AI,” an energy pitch (“plug, baby, plug!”), and a push to soften the EU AI Act to help startups compete.
At the New Delhi AI Impact Summit (February 2026), the tone shifted dramatically. Macron pivoted to ethics and shared sovereignty: international cooperation, protection of minors, algorithmic transparency, and a direct warning against the concentration of technological power—with an unmistakable jab at Elon Musk’s Grok.
The France-India coalition, the emphasis on youth as a political subject, and the call for global governance mark the main divergence in tone and substance between the two summits. In Paris, Macron was selling; in New Delhi, he was governing.
And predictably, this gave ammunition to the Americans. The White House’s AI guru, Sriram Krishnan, went hard on Europe, accusing the continent of “too much doomerism” and “too little innovation.” At Davos, White House science adviser Michael Kratsios called the EU AI Act “an absolute disaster.”
And yet... $3 billion poured into European AI startups in January 2026 alone, with at least 10 new unicorns. France leads the pack with Mistral AI, Poolside, Homa, Photoroom, and the newly emerged Ineffable Intelligence. Germany counters with Black Forest Labs and Nscale. The money is here. The talent is here.
More importantly, the infrastructure is finally being built. This week: ASTRA, a consortium of six European PE funds managing €4 billion, was formed to coordinate sovereign tech investment. The ÆTHER consortium entered exclusive negotiations for AI gigafactories. Spain announced a €1.5 billion Deep Tech strategy. The European Parliament disabled AI features on lawmakers’ devices over data sovereignty fears. And EuroHPC launched its Frontier AI Grand Challenge to train the first truly European frontier model.
Meanwhile, the biggest brain in European AI, Yann LeCun, officially launched AMI Labs from Paris—a €3 billion bet that LLMs are a “dead end” and that “world models” are the future. His parting shot at Meta? “You certainly don’t tell a researcher like me what to do.”
One email a week to understand the latest advancements in AI for Consultant, Market and Competitive Intelligence professionals.
Don’t forget to share it!!! Cheeerrrssss :-)
Macron in New Delhi: From Business Pitch to Global Governance
As the AI Impact Summit shifts to the Global South, Macron pivots from selling French AI to governing its global risks—calling for shared sovereignty, child protection, and algorithmic transparency.
The contrast between Macron’s two major AI speeches could not be starker. At Paris in 2025, the French President announced €109 billion in pledges from MGX, BlackRock, Amazon, Microsoft, and Mistral, positioning France as the “world’s No. 3” in AI behind the US and China. The focus was squarely on attracting capital, simplifying regulation, and leveraging France’s clean energy advantage.
In New Delhi, the narrative shifted to international cooperation and ethical governance. Macron warned against technological dependency on the US or China, called for robust algorithmic transparency, and made the protection of minors a central policy pillar.
The France-India strategic coalition on AI and the emphasis on youth as a political subject mark a significant departure. For business leaders, this signals that France is playing two games simultaneously: attracting investment at home while building a regulatory coalition abroad.
Sources:
Don’t want to be dependent on US or China: Macron on AI innovation
India’s AI summit draws global leaders, big pledges and some chaos
India’s $200 Billion AI Ambition: The Global South Takes Center Stage
With $210 billion pledged by Reliance and Adani, and partnerships signed with OpenAI, Anthropic, and Infosys, India’s AI Impact Summit signals the rise of a new sovereign AI power.
The India AI Impact Summit 2026—held in New Delhi from February 19-20—marked a watershed moment: the first global AI summit hosted by a Global South nation. Inaugurated by Prime Minister Narendra Modi, it drew delegations from over 100 countries, including 20 heads of state and 300,000 participants.
The numbers are staggering. India’s two biggest conglomerates, Reliance and Adani, pledged a combined $210 billion in domestic AI and data infrastructure. OpenAI signed a partnership with the Tata Group, while Anthropic partnered with Infosys and opened an office in Bangalore. The Indian government aims to attract $200 billion in AI investment over the next two years.
For intelligence professionals, the message is clear: India is no longer a backend office. It is positioning itself as a sovereign AI hub with the scale, talent, and capital to set its own rules.
Source: Chaos, confusion and $200 billion dreams: What I saw at India’s AI summit
The White House vs. Europe: “Too Much Doomerism, Too Little Innovation”
US officials escalate their criticism of the EU’s regulatory approach, calling the AI Act “an absolute disaster” and accusing Europe of stifling the very innovation it claims to champion.
The transatlantic AI divide deepened this week. Sriram Krishnan, Senior White House Policy Adviser on AI, publicly urged the EU to drop its “AI doomerism”—a mindset he argues focuses excessively on risks at the expense of innovation. At Davos, Michael Kratsios, White House science and tech adviser, was even more blunt, calling the EU AI Act “an absolute disaster.”
This rhetoric is strategic: Washington is positioning deregulation as a competitive weapon, arguing that Europe’s rigid frameworks push startups and investors to more permissive markets. For European business leaders, this creates a double bind: comply with the world’s strictest AI regulation, or risk losing talent and capital to the US.
Yet the data tells a more nuanced story. With $3 billion invested in European AI startups in January alone and 10+ new unicorns, the continent is proving that regulation and innovation can coexist—if the capital follows.
Sources:
White House official says EU must abandon AI ‘doomerism’
Vita husets AI-guru går hårt åt Europ
a
Europe’s $3 Billion January: 10 New AI Unicorns Defy the “Innovation Gap” Narrative
Despite Washington’s criticism, European AI startups raised approximately $3 billion in January 2026 alone, with France leading the charge in what may be the continent’s most productive funding month ever.
The numbers speak louder than the rhetoric. In January 2026 alone, approximately $3 billion flowed into European tech startups, with AI capturing the lion’s share. At least 10 startups crossed or exceeded the billion-dollar valuation mark:
ElevenLabs (UK) — voice AI
Ineffable Intelligence (France) — frontier AI
Mistral AI (France) — foundation models
Poolside (France) — code generation
Homa (France) — gaming AI
Photoroom (France) — image editing AI
Black Forest Labs (Germany) — Flux image generation
Nscale (Germany) — AI infrastructure
Humans& ($480M funding in January 2026)
This excludes February rounds for ElevenLabs and Ineffable Intelligence. For most companies listed, recent funding rounds fall in the $100M–$500M range, confirming Europe’s ability to produce capital-intensive AI champions despite regulatory headwinds.
ASTRA: Europe’s €4 Billion Private Equity Answer to the Sovereign Tech Gap
Six European VC funds unite under the ASTRA consortium, pooling €4 billion in assets to coordinate investment across AI, quantum, robotics, and critical technologies.
ASTRA is a new European private equity consortium launched in early 2026, uniting six venture capital funds: 360 Capital (France-Italy), Bullhound Capital, Redstone, Truffle Capital, UVC Partners, and 28Digital. Together, these funds manage over €4 billion, count more than 800 investments, and operate from 19 offices across nearly every European country.
The consortium’s objective is to dynamize investment in the European tech ecosystem by mutualizing expertise, deal flow, and resources. ASTRA will coordinate investments across AI, robotics, quantum computing, energy storage, and the critical technologies at the heart of Europe’s sovereignty ambitions.
For intelligence professionals, this signals a structural shift: European capital is no longer fragmented but increasingly coordinated to match the scale of US and Asian competitors.
Source: ASTRA: il nuovo consorzio dei private equity
ÆTHER Consortium: Building Europe’s AI Gigafactories
The ÆTHER infrastructure consortium enters exclusive negotiations for a strategic industrial site dedicated to AI gigafactories, backed by an 80/20 debt-equity model.
The ÆTHER infrastructure consortium has initiated exclusive negotiations for the establishment of a strategic industrial site dedicated to AI gigafactories in Europe. This initiative is aligned with the EU’s AI Continent Action Plan and the broader European investment objectives for artificial intelligence.
The consortium operates on a mixed financing model: 80% debt and 20% equity from consortium members. The first operational deployment was announced in February 2026. For business leaders in regulated sectors, ÆTHER represents a credible path to indigenous European AI infrastructure—operated by European partners, for European stakeholders.
Spain’s €1.5 Billion Deep Tech Gamble: A New Sovereign AI Player Emerges
Spain launches one of Europe’s most ambitious national Deep Tech strategies, committing €1.5 billion to AI and over €800 million to quantum computing.
Spain is preparing to launch its national Deep Tech strategy, with an initial allocation of €1.5 billion for AI and more than €800 million for quantum computing. This represents one of the most ambitious European investments in sovereign AI and related infrastructure to date.
For market intelligence professionals, Spain’s entry into the sovereign AI race adds a significant new player to the European ecosystem, potentially reshaping the competitive dynamics between France, Germany, and the Nordics.
Source: What are deep technologies or deep tech?
EuroHPC’s Frontier AI Grand Challenge: Europe’s Moonshot for a Sovereign Foundation Model
The European Commission and EuroHPC launch a flagship competition to train the first European frontier AI model with 400+ billion parameters on sovereign supercomputers.
The Frontier AI Grand Challenge, launched on February 13, 2026, is a flagship EU-wide competition to close the strategic gap in high-end AI. The challenge invites European teams to develop frontier AI models with a computational capacity equivalent to at least 400 billion parameters, using efficient architectures like Mixture-of-Experts (MoE).
The winning project will receive up to 2.5% of EuroHPC’s total computing capacity for one year, including access to JUPITER (Germany’s exascale supercomputer) and Alice Recoque (France’s upcoming exascale system). The resulting models will be open and available to public authorities, scientific communities, and businesses across Europe.
EuroHPC acknowledges the stakes bluntly: “no frontier-level models are currently being operated at a scale comparable to the leading global actors.” The submission deadline is April 13, 2026.
Source: Frontier AI Grand Challenge Puts European HPC Sovereignty in the Spotlight
Mistral AI Acquires Koyeb: Completing the Sovereign Stack
Following its €1.2 billion Swedish data center investment, Mistral AI’s acquisition of cloud startup Koyeb signals a decisive vertical integration play—from model weights to serverless infrastructure.
Mistral AI has acquired French cloud startup Koyeb in its first M&A deal, marking a strategic pivot toward end-to-end sovereign AI infrastructure. This move ensures Europe’s capacity to control not only AI models but also the underlying cloud platforms, shifting the narrative from dependency on Silicon Valley to homegrown capability.
Combined with its €1.2 billion commitment to data centers in Sweden (which we discussed in Week 7), Mistral is building a fully European-controlled stack—from model weights to physical servers to the serverless deployment layer. The acquisition facilitates industrial, governmental, and regulated sector deployments where full-stack sovereignty is a prerequisite.
For competitive intelligence professionals, this vertical integration is a textbook play: owning every layer reduces dependency and creates a defensible moat.
Source: Mistral AI buys cloud startup Koyeb
Policloud’s 1,000 Sovereign Micro-Data Centers: Decentralizing AI Infrastructure
At WAICF 2026, French startup Policloud announces plans for 1,000 sovereign AI micro-data centers by 2030, offering a distributed alternative to centralized hyperscaler models.
At the World AI Cannes Festival (WAICF) 2026, French company Policloud announced plans to accelerate its international expansion by deploying 1,000 sovereign AI micro-data centers by 2030. This distributed approach offers an alternative to the centralized hyperscaler model, bringing compute closer to the edge while maintaining European data sovereignty.
For regulated sectors and public institutions, distributed micro-data centers provide a compelling combination: lower latency, local legal jurisdiction, and reduced concentration risk.
Europe’s Vertical AI Strategy: Playing to Sectoral Strengths
European leaders and policymakers argue the continent should deploy sovereign AI within its sectors of excellence—pharma, manufacturing, finance—rather than trying to replicate the American model.
A growing consensus among European AI leaders and policymakers is emerging: rather than attempting to build general-purpose AI champions to rival OpenAI or Google, Europe should focus on deploying sovereign AI within its sectors of industrial excellence—pharmaceuticals, advanced manufacturing, and financial services.
This “vertical” specialization is increasingly seen as Europe’s best strategic asset for differentiation and relevance. By combining domain expertise, high-quality data, and regulatory alignment, European companies can carve out defensible niches that American generalist models cannot easily penetrate.
Source: Europa’s gevecht om AI-relevantie
Yann LeCun Launches AMI Labs: A €3 Billion Bet That LLMs Are a Dead End
Meta’s former Chief AI Scientist bets his reputation—and €500 million in initial funding—on “world models” as the true path to superintelligence, headquartered in Paris.
Yann LeCun, the Turing Award winner and Meta’s Chief AI Scientist for over a decade, has officially left Meta to launch AMI Labs (Advanced Machine Intelligence)—pronounced “ami,” meaning “friend” in French. The startup is headquartered in Paris, with LeCun as executive chairman and Alexandre LeBrun (founder of health-tech firm Nabla) as CEO.
AMI Labs is raising €500 million at a €3 billion valuation, backed by the conviction that LLMs are a “dead end” on the road to superintelligence. Instead, LeCun is betting on world models—AI systems that build abstract representations of reality using Joint Embedding Predictive Architecture (JEPA), rather than generating predictions token by token.
His departure from Meta was far from quiet. After Mark Zuckerberg hired Alexandr Wang (Scale AI founder) to lead Meta’s new Superintelligence Labs—making Wang effectively LeCun’s boss—LeCun pushed back. He called Wang “inexperienced” and admitted that Llama 4 benchmarks were “fudged.”
For the European AI ecosystem, AMI Labs is a massive signal. LeCun himself noted: “There is a huge demand from the industry and governments for a credible frontier AI company that is neither Chinese nor American.” Meta will be a partner—but not a funder—of the new venture.
Sources:
Yann LeCun’s new venture is a contrarian bet against large language models (MIT Technology Review)
“You certainly don’t tell a researcher like me what to do” says LeCun as he exits Meta
Who’s behind AMI Labs, Yann LeCun’s ‘world model’ startup (TechCrunch)
Thanks for reading The C&M Intelligence Leader’s Wine! Subscribe for free to receive new posts and support my work.