🌊 Where AI Capital Is Consolidating
Where capital flows. How the best build.
👋 Hey, I’m Ivan. I study where capital flows and how top founders build. Today’s Capital Flow drop tl;dr:
🌊 The first wave of AI markets has consolidated
🌊 Q2's mega AI rounds: 16 startups raised $31B
🌊 80+ startups became unicorns in H1 2025 (15 in Europe)
🌊 European AI Series A heatmap
🌊 The GTM AI startup wave (YC signals)
🌊 Consumer AI: emotion, memory, and intimacy
🌊 Bonus: Spain’s €10M+ rounds
🌊 Things I’ve enjoyed reading in AI this week
🌊 The first wave of AI markets has consolidated
“You can now name (some of) the winners.” – Elad Gil
What Happened
Elad Gil (invested in 40 unicorns) wrote a great piece on how, after years of noise, the first wave of AI markets has solidified. We now have clear(er) leaders across foundational AI categories.
The Details
Mostly locked-in markets (winners known):
LLMs → OpenAI, Anthropic, Meta, Mistral, Google
Code → Copilot, Cursor, Claude Code, Devin
Legal → Harvey, CaseText
Medical Scribing → Abridge, Ambience
Customer Service → Decagon, Sierra
Search → Perplexity, OpenAI, Google
Markets still forming:
Sales agents
Compliance
Accounting
Security
Agent infra
AI-driven roll-ups
So What
Elad often bets on markets > founders (afaik and researched). With that in mind:
If you’re in a “crystalised” market: build a wedge or GTM advantage.
If you’re in a forming market: now’s the time to claim the category.
Next waves will unlock as models hit new fidelity thresholds (e.g. GPT-5, Claude 3.5).
🌊 Q2's Mega AI Rounds: 16 Startups, $31B Raised
What Happened:
In Q2 2025, 16 startups raised $500M+ rounds, totalling over $31B in capital. AI continues to dominate, with 12 of the 16 are AI-native or adjacent.
The Details
Here’s a simplified leaderboard of the biggest raises (credit to
So what?
AI is no longer early-stage hype, it’s dominating later stage rounds.
$2B seed rounds (e.g. Thinking Machines Lab) are a signal a new elite tier of “startup.” (or signs of a bubble, or both). We’ve talked about before about how this AI cycle is probably simultaneously over and under-hyped (short-run overrated, medium-term underrated).
Don’t compare your $2M raise to these outliers, these are outliers.
Instead, study what makes them investable at this scale: strategic positioning, infra depth, and geopolitical alignment
🌊 80+ Unicorns in H1 2025 (Europe Holds Its Own)
What Happened
80+ startups became unicorns in H1 2025. 15 of them were European across AI, quantum, defense, and health. It’s still a U.S.-led wave (52 unicorns), but Europe is holding its own, especially in deep tech.
The Details
1) Tide (🇬🇧) – Business banking for SMEs
2) MUBI (🇬🇧) – Global streaming for indie films
3) Oxford Ionics (🇬🇧) – Scalable quantum computing
4) Verdiva Bio (🇬🇧) – Sustainable plant-based meat
5) Isomorphic Labs (🇬🇧) – AI drug discovery
6) PhysicsX (🇬🇧) – AI-powered industrial design
7) Lovable (🇸🇪) – AI platform, build apps without coding
8) Neko Health (🇸🇪) – scanning for preventive healthcare
9) Parloa (🇩🇪) – conversational AI platform
10) Quantum Systems (🇩🇪) – Drones for defense
11) Zama (🇫🇷 / 🇨🇭) – Confidential blockchain protocol
12) Tines (🇮🇪) – No-code security workflows
13) Namirial (🇮🇹) – techn for digital transactions
14) Diagnostyka (🇵🇱) – medical diagnostic laboratory
15) TEKEVER (🇵🇹) – Drones for maritime patrol
The full map:
So what?
Europe’s unicorn playbook is shifting:
AI-native platforms are scaling: Isomorphic, Lovable, PhysicsX
Quantum & deep tech are investable again: Oxford Ionics, Zama
Defense is real: Quantum Systems, Tekever, public sector buyers are back
Healthcare on the rise: Neko, Diagnostyka + surf the aging population wave
🌊 European AI Series A Heatmap
What happened
30 European AI startups raised €10M+ Series A rounds in H1 2025—totalling $957M in capital deployed.
The Details
AppliedAI – AI + Insurance (London) – $55M
Tandem Health – AI + Health (Stockholm) – $50M
Quibim – AI + Biotech (Valencia) – $50M
NexGen Cloud– AI + GPU infra (London) – $45M
Samaya AI – AI + Research tooling (London) – $43M
Bioptimus – AI + Biotech (Paris) – $41M
Latent Labs – AI + Biotech (London) – $40M
Relay – AI + E-commerce (London) – $35M
cartagon.ai – High performance GPUs – $32M
Salience Labs – AI + Chips infra (Oxford) – $30M
Unique – AI + Finance (Zürich) – $30M
Sereact – AI + Robotics (Stuttgart) – $25M
Wordsmith AI – AI + Legal Tech (Edinburgh) – $25M
Maze – AI + Cloud security (London) – $25M
Apheris – Model performance (Berlin) – $20.8M
Synthflow AI – AI + Voice Interfaces (Berlin) – $20M
GetWhy – AI + Research (Copenhagen) – $17M
ecoplanet – Energy Management (Munich) – $16M
Validation Cloud – AI + Web3 Infra (Zug) – $15M
TurinTech AI – Code Optimisation (London) – $15M
Gradient Labs – AI + CS (London) – $11.8M
Aive – AI + Video Automation (Levallois) – $12M
Stotles – AI + Public Sector Sales (London) – $10M
Inven – AI + B2B M&A SaaS (Helsinki) – $11.2M
Yaspa – AI + Fintech (London) – $12M
Jua.ai – AI + Weather Forecasting (Zürich) – $10M
ocell – AI + Climate Intelligence (Munich) – $10M
Flank – AI + Enterprise Governance (Berlin) – $10M
tldraw – AI + Developer Tooling (London) – $10M
So what?
Three shifts stand out:
Vertical AI is dominant: 70% of rounds are in applied sectors like biotech, insurance, infra, and voice.
London is the epicenter: 11 of the 30 rounds are London-based, especially infra and vertical AI tooling.
Biotech & Infra lead in check size: 6 out of the top 10 rounds are in biotech or GPU infra.
🌊 The GTM AI Startup Wave (YC Signals)
What happened:
Over the past 18 months or so we’re seeing a big GTM AI wave forming, best represented by how many of these players are entering Y Combinator in recent batches.
The details:
🔒 *The rest of this edition is for Startup Riders Pro readers only.*