Illustration for: September's Megadeals Already Top $8 Billion

September's Megadeals Already Top $8 Billion

Six companies -- The Boring Company, Cognition, Positron AI, Mach Industries, Harvey, and Ayar Labs -- have absorbed more than $8 billion in the first ten days of September, almost all of it into physical AI infrastructure and applied AI.

By the Numbers

$3.0B @ $23B
Boring Company
$2.0B @ $48B
Cognition
$875M @ $5B
Positron AI
$550M @ $15.5B
Harvey
$600M @ $3.7B
Mach Industries
TC
By the Funding Desk
Edited by Trace Cohen · Early-stage VC & angel · Founder, New York Venture Partners
1 min read
ShareXLinkedInEmail
TC

The VC Read · Trace's Take

Trace Cohen

The concentration number that should worry LPs isn't the $8B total, it's that six names account for nearly all of it while seed and Series A activity outside AI infrastructure stays comparatively flat. When a market gets this concentrated in physical-AI mega-rounds, the funds that missed this specific vintage of six companies are structurally behind for the cycle, not just for the quarter.

Analysis

Six companies have absorbed more than $8 billion in disclosed venture and growth capital in the first ten days of September:

  • [The Boring Company](/pulse/company/the-boring-company) -- $3.0B Series D at a $23B valuation.
  • Cognition -- $2.0B round at a $48B valuation.
  • Positron AI -- $875M Series C at a $5B valuation.
  • Mach Industries -- $600M Series C extension at a $3.7B valuation.
  • Harvey -- $550M round at a $15.5B valuation.
  • Ayar Labs -- $650M cumulative across 2026, now valued above $5B.

Tech Startups' own tally of that week's ten largest disclosed financings put roughly 93% of the week's $4.81 billion into just three of those names -- Boring Company, Positron AI, and Mach Industries -- a reminder of how concentrated even a broad-looking funding week actually is once you sort by dollar size rather than deal count.

- The Boring Company -- $3.0B Series D at a $23B valuation.

What connects five of the six: every one of them sells into a physical or infrastructure bottleneck rather than a software layer that could theoretically be replicated by a well-funded competitor in months. Tunnels, inference chips, autonomous manufacturing, legal AI with enterprise lock-in, and optical interconnects all share one trait -- they take years and capital-intensive buildouts to replicate, which is exactly the kind of moat growth investors are currently willing to pay a premium for.

Harvey is the outlier in that list -- a legal AI company competing in an application layer where model access, not physical infrastructure, is the primary constraint. Its $15.5 billion valuation on $550 million raised is a bet that enterprise legal workflows and data lock-in create a moat even without the hard-asset backing the other five names share.

For LPs sizing exposure to this vintage: six deals is not a diversified thesis, it's a bet on physical-AI infrastructure holding its value through a hardware refresh cycle most of these valuations assume won't happen for several years.

ShareXLinkedInEmail

Key Sources

2 sources

THE WIRE in your inbox— Tech, startup & VC news with Trace's take. Free, no spam.