Almost 90 new unicorns have been minted so far this year — here they are

TechCrunch (via Crunchbase/PitchBook) identified a wave of VC-backed unicorns in 2026, most AI-related, including Promethus at $41B valuation, Recursive at $4.65B, and Nextop AI at $4.2B. The list also spans fintech (April Rogo at ~$2B; Slash at ~$1.4B), healthcare (MiRus at $4.41B; Stipple Bio at $2.25B), and select crypto/security plays (TRM Labs at ~$1B; Rain at ~$1.9B), signaling continued investor risk appetite across venture markets.

Analysis

The public-market read-through is less about the unicorn tally itself and more about the funding stack underneath it. A dense crop of AI-native startups implies continued demand for compute, cloud, networking, security, and model-distribution layers, which is structurally supportive for NVDA, GOOGL, and AMZN over the next 6-18 months. The second-order winner is not the app layer; it is the infrastructure toll collectors that get paid regardless of which startup survives.

The more interesting risk is competitive abrasion for incumbent software and workflow platforms. When private companies are being funded at premium marks to automate coding, support, payments, analytics, and healthcare admin, public SaaS names like CRM and payments/search-adjacent platforms such as PYPL face a gradual multiple headwind as investors question long-term seat growth and pricing power. That effect will not show up in next quarter revenue, but it can compress valuation over 1-3 quarters if AI-native product demos start converting into enterprise pilots.

Private-markets and deal finance should see a short-term uplift in fees and transaction activity, which is modestly constructive for BX, BLK, and JPM. The contrarian issue is exit quality: if the IPO window stays selective, these unicorns become a mark-to-model rather than monetization story, and that eventually feeds back into venture fundraising discipline. Falsifiers for the bullish infrastructure thesis are a hyperscaler capex slowdown, a turn lower in Nvidia order visibility, or a sudden rise in software churn as buyers delay AI spend.

Key Decisions for Investors

• Long NVDA on 5-10% pullbacks over the next 1-3 months; thesis is continued AI capex breadth, with upside if hyperscaler spend stays above consensus. Falsify on weaker data-center demand commentary or order delays.
• Pair trade: long GOOGL / short CRM for 1-3 months to express AI-native workflow substitution risk without taking broad beta. Risk/reward improves if enterprise software billings weaken faster than cloud demand.
• Overweight BX and BLK on a 6-12 month horizon as proxies for private-market fee flow and eventual exit optionality. Thesis breaks if the IPO window remains shut and unrealized marks stop moving.
• Use JPM as a relative long versus PYPL if you want a cleaner monetization of private-company financing and venture deal flow; avoid overpaying for payments incumbents that may face AI-driven workflow disintermediation.