AI & Intelligence Briefing — June 23, 2026
The EU picks a winner to build its own 400-billion-parameter open model as Brussels finalises the AI Act Omnibus and a viral doomsday scenario reignites the sovereignty debate.
Read briefingIntelligence Briefings
Following European AI research, deployment, governance, and infrastructure: what's being built, what's working, and what isn't.
All briefingsEurope's approach to artificial intelligence is not primarily a technology story. It is a governance story: who decides which risks are acceptable, which applications are permitted, and where the infrastructure runs.
The AI & Intelligence briefings track the structural decisions beneath the headline announcements. The regulatory mechanics. The capital movements. The infrastructure investments that will determine what European AI actually looks like in five years.
The AI Act is the world's first comprehensive AI regulation, and it is already reshaping product development, procurement, and investment decisions across Europe.
The briefings track the practical implementation. Not the political narrative; the technical annexes, the compliance timelines, the delegated acts, and the interpretations that determine what the law actually requires of builders, deployers, and public institutions.
Particular attention goes to the high-risk classification system, the GPAI provisions covering foundation models, and the enforcement architecture being built by national market surveillance authorities.
European AI sovereignty is not decided at the model layer. It is decided at the infrastructure layer.
The briefings follow where compute is being built, who owns it, and on what terms. This includes the European AI factories programme, national sovereign cloud initiatives, and the Mistral and Aleph Alpha model investments.
The core question is simpler and harder than it looks: will European institutions run on European infrastructure, or remain dependent on US hyperscalers? These decisions are being made now. They will be difficult to reverse.
Europe's AI research base is significant and underreported.
The briefings cover frontier model development: the major lab funding rounds, the academic-to-commercial transitions, the Horizon Europe research investments. They situate these inside the broader question of where European scientific capacity is being directed.
The interest is less in which model benchmarks best, and more in what kind of institutional structure is being built to sustain European AI research over the long term.
The EU picks a winner to build its own 400-billion-parameter open model as Brussels finalises the AI Act Omnibus and a viral doomsday scenario reignites the sovereignty debate.
Read briefingMistral eyes a €3 billion raise at €20 billion valuation as DeepMind and Nebius back European physical AI — and Brussels closes in with operational labelling rules.
Read briefingBrussels tables its tech sovereignty package as the EU flagship AI gigafactory plan stalls; PhysicsX raises $300M for physics AI; mathematicians issue the Leiden Declaration against training on their work.
Read briefingSPRIND closes its €125M next-paradigm AI lab challenge; Mistral's first sovereign data centre goes live near Paris; Malta becomes the first country to give every citizen free ChatGPT Plus.
Read briefingEU's Cloud and AI Development Act lands tomorrow; Recursive Superintelligence raises $650M from London; Mistral acquires Vienna's physics AI startup Emmi.
Read briefingEU legislators grant 16 more months for high-risk AI compliance, SAP acquires Freiburg's Prior Labs for €1B+, and Cohere merges with Aleph Alpha to build a transatlantic sovereign AI challenger.
Read briefing