The recent funding log points to a single pattern: large checks are still being written, but they are going to a narrower band of companies that can absorb enormous amounts of capital before proving a conventional business model. BrightAI's $97 million Series A, Science's $400 million round and Anthropic's Nasdaq listing are not the same kind of event, yet each shows where US venture dollars are concentrating in late 2026. The common thread is scale, and the willingness of investors to fund it early.
Hardware and Infrastructure Take the Big Checks
Two of the three logged stories involve companies whose products sit in the physical world. BrightAI Inc., which helps industrial companies detect malfunctioning equipment, raised $97 million in a Series A led by Smith Point Capital, as SiliconANGLE reported. The round drew other venture firms and GE Vernova Inc., an energy equipment maker. The size of a Series A in industrial monitoring is notable on its own; it suggests investors are pricing the deployment problem, not just the software. Sensors, installations and integration with existing plant equipment cost money before any subscription revenue arrives.
The same logic appears, larger, in Crunchbase News' weekly roundup, where brain-computer interface company Science led U.S. startups with a fresh $400 million. That is a hardware and clinical-research business by nature, and the figure implies investors are willing to fund years of development before commercial scale. Both deals point away from the cheap-to-launch software model that defined the previous funding cycle and toward companies that need capital to build something physical.
AI's Capital Question Moves to the Public Market
Anthropic's situation, discussed by The Verge in the third episode of its AI in real life series, is different in structure but similar in implication. The AI lab behind Claude went public on the Nasdaq the previous Thursday, as The Verge noted, and the conversation with senior AI reporter Hayden Field focused on whether the company can afford to put safety first. The relevant fact for this beat is not the size of the listing but the shift in where the capital comes from. A private AI lab raising private rounds answers to a small group of investors; a public one answers to a market that prices quarterly expectations.
The Verge framing, that the numbers are gigantic but not really that important for the conversation, is the analyst's point too. What matters is that the funding question for the largest AI developers has moved from "who will write the check" to "what does the check require." That is a different kind of pressure, and it lands on safety spending, staffing and research timelines.
The Middle of the Market Looks Thin
Read together, the three stories describe a barbell. At one end sit very large rounds for capital-intensive hardware and AI; at the other, early-stage companies still raising modest seed and Series A checks. What the log does not show is much activity in the middle, the $20 million to $60 million expansion rounds that once carried software companies from product to national sales. BrightAI's $97 million is large for a Series A and would once have been a growth round. Science's $400 million would have been a late-stage figure. The thresholds have moved.
For US technology companies, that has a practical consequence. A startup with steady revenue but no hardware and no AI story may find its next round harder to price than a pre-revenue company with a factory floor or a model to train. Founders in that position face a choice between stretching runway and accepting terms they would previously have refused.




