The funding cycle has turned from rewarding ambition to pricing it. Across dealmaking, research spending and corporate cost structures, the companies and backers that spent the boom years selling a story are now being asked to show the economics. For US technology companies, that means valuations, headcount and even research budgets are increasingly set by what can be demonstrated rather than what can be imagined.
The acquirer's market
The clearest sign is in exits. As Crunchbase News reported, a number of companies whose names might feature on a playlist of 2021 boom-era unicorns have sold to acquirers in recent months, including one-time highfliers like Airtable, Headspace and Miro. In many cases, the companies are selling at valuations which, while not puny, are still far below their former peaks.
That is the defining feature of the current funding environment. The buyers are not paying for the peak narrative; they are paying for what the asset can do now. A sale is still a sale, and the prices are not trivial, but the distance between the boom-era mark and the acquisition price is where the story lives. For founders and their backers, the practical consequence is that the exit route has narrowed: the strategic buyer with a balance sheet is replacing the late-stage crossover round as the most plausible path for a company whose growth no longer matches its last valuation.
This matters for the US market specifically because so much of the past decade's venture activity was built on the assumption that a private mark could be defended until a public listing or a larger round arrived. When the acquirers set the price, that assumption breaks. Employees holding equity learn what the new mark is. Funds that marked positions at the old peak face the write-down. And the next cohort of founders hears the message that a durable business at a modest valuation is worth more than a spectacular one that cannot raise again.
AI as a research line item, not a valuation story
The second thread is about what investment in artificial intelligence is actually buying. As The Verge reported, an investment research firm wanted to prove AI could revolutionize drug development, so it teamed up with Anthropic to see whether an unreleased version of Claude could make scorpion antivenom. It could not. It came up with some promising starting points, but nothing close to an actual drug, experts told The Verge.
For the funding beat, the detail worth noting is who was doing the spending. Citrini Research, an investment research firm, funded an experiment designed to test a thesis rather than to launch a product. That is a different kind of AI money from the venture rounds that dominated the boom. It is diligence spending, thesis testing, the cost of finding out whether a technology claim survives contact with experts. The result was neither a triumph nor a debunking: promising starting points, no drug.
That is a useful calibration for US investors. The question is no longer whether AI can be applied to a sector, but what a given application actually produces, and on what timeline. Capital that once flowed to anyone with a plausible AI story is now being spent to check whether the story holds. The firms writing those checks are still writing them, but they are buying evidence, not adjacency.
