Capital concentration in the AI sector is shattering historical records. According to the latest data, 80% of global venture capital in Q1 2026 flowed to AI, and just four companies swallowed 65% of the funds.
Latest Valuation Rankings (May 2026)
| Rank | Company | Valuation | Notes |
|---|---|---|---|
| 1 | Anthropic | $900B+ | Surpassed OpenAI to take #1 |
| 2 | OpenAI | $852B | Driven by GPT-5 expectations |
| 3 | xAI | $250B | Grok series continues iterating |
| 4 | Databricks | ~$134B | Data + AI platform |
| 5 | Waymo | $126B | Autonomous driving leader |
| 6 | DeepSeek | $45B | Only Chinese AI company in top ten |
| 7 | Figure AI | $39B | Humanoid robots |
| 8 | SSI | $32B | Space technology |
Three Key Signals
1. Anthropic Overtakes OpenAI
Anthropic surpassed OpenAI for the first time with a $900B valuation versus $852B. Behind this overtake are several factors:
- Claude series models’ penetration in the developer market continues to climb
- Commercialization of vertical product lines like Claude Design
- Enterprise customer acceptance of the “safe AI” narrative
- Investor concerns over OpenAI’s governance controversies
2. The Significance of DeepSeek’s $45B
DeepSeek is currently the only Chinese AI company in the global top eight. While the $45B valuation shows a significant gap with the top two, the valuation growth rate is quite aggressive given its founding time and fundraising scale. Its differentiation strategy:
- Extreme cost-effectiveness: providing comparable performance at far lower inference costs than GPT-4
- Open-source strategy: V3/R1 series models open-sourced, building developer ecosystem
- V4 expectations: Next-gen model adapts to NVIDIA Blackwell architecture, costs dropping another 20x
3. Capital Concentration Hits New Highs
In Q1 2026’s $300B global venture capital, 80% went to AI, with four companies taking 65%. This means:
- The fundraising environment for mid-tier and small AI startups has deteriorated sharply
- Capital is accelerating toward the top
- “AI infrastructure” has become a capital consensus direction
Infrastructure Arms Race Escalating in Parallel
The five major tech giants (Amazon, Google, Meta, Microsoft, Oracle) are expected to spend $830B on AI infrastructure capex in 2026:
| Year | Capex |
|---|---|
| 2024 | $260B |
| 2025 | $435B |
| 2026 | $830B (expected) |
| 2027 | $1.15T (expected) |
Alphabet recently issued six-tranche euro bonds raising at least €3B, directly fueling the AI race. This is not routine financing—it’s the market reminding everyone that the AI arms race is evolving into a debt race.
Landscape Assessment and Action Advice
| Role | Recommendation |
|---|---|
| Investors | Head concentration is already extremely high; focus on vertical application layers and on-device model opportunities |
| Entrepreneurs | The general model race window has closed; differentiated scenarios are the only way out |
| Enterprise users | Multi-model strategy (Anthropic + DeepSeek + open source) to diversify vendor lock-in risk |