PitchBook

86% of U.S. venture funding in H1 2026 went to AI startups, according to PitchBook, making its data a key gauge of a sharply concentrated private-market boom.

Who they are

PitchBook appears in coverage as a private-market data source: reporters use its funding, valuation, investor and exit figures to put startup financings, acquisitions and public-market milestones in context, from Robinhood’s IPO to Amazon’s acquisition of robotics startup Rivr.

The recent arc

Coverage accelerated through 2026, peaking in Q3, as PitchBook’s figures became evidence for the scale and composition of the AI-led venture cycle. Its H1 2026 data put U.S. venture funding at $412.7 billion, with AI startups accounting for $355.9 billion, while it counted seven U.S. rounds above $1 billion in Q2. Recent Financial Times coverage extended that lens to robotics and physical AI, with roughly $48 billion raised year to date, and to quantum computing, where VC investment exceeded $4 billion year to date.

The latest stories shift from broad funding totals toward the consequences of concentrated capital and fast-rising valuations. Axios cited PitchBook in reporting that 42 startups had become decacorns in 2026 and that the average time to a $10 billion valuation fell to 5.5 years in 2025 from 9.4 years in 2020. At the deal level, its historical valuations have also been a recurring benchmark in acquisitions including Superhuman’s purchase of GPTZero, Palo Alto Networks’ purchase of Console, and Amazon’s purchase of Rivr.

The tension

The coverage centers on an uneven private-market recovery: huge AI, robotics and quantum rounds and faster paths to decacorn status sit alongside PitchBook’s finding that half of U.S. unicorns have not raised funding in three years. In effect, the data depicts VCs concentrating capital in a small set of technology themes and large rounds while many earlier high-valuation companies face a far less liquid financing environment.

Why it matters

If this pattern persists, PitchBook’s measures will remain central to judging whether the AI investment surge is broadening into durable venture-market renewal or merely compressing capital, valuations and exits into a small group of winners. The gap between new decacorns and inactive unicorns also makes its funding and valuation benchmarks consequential for investors, acquirers and companies trying to assess what private-market prices signal about eventual exits.

Related: AI · VCs · IPO · Superhuman acquires AI detection startup GPTZero, which has 19M+ regis · Sources: Palo Alto Networks paid $500M in cash and stock to acquire ag