The funding market in late 2026 is not broad-based. It is concentrating. Three stories logged on this beat tell the same story from different angles: Mercor raising $750 million at a $20 billion valuation for AI agent evaluation, Crunchbase recording a record 1,900-company Unicorn Board in October with 34 new entrants, and Anthropic seeking up to $88.5 billion tied to a public market debut. The common thread is that capital is clustering around AI infrastructure and the companies that supply it, while the wider unicorn population is being repopulated by younger, faster-scaling firms.
The AI layer is where the money is
Mercor.io Corp. is a training data company. It sells human expertise - drawn from a network of more than 5 million domain experts - to train and evaluate AI models, and it is now pushing that evaluation work into enterprises, as SiliconANGLE reported. A $750 million raise at a $20 billion valuation for a company in that position says something specific: investors are not only paying for model builders, they are paying for the supply chain that makes models usable and trustworthy. Evaluation is the part of the AI stack that sits between a model and a paying customer, and the market is pricing it as strategic infrastructure rather than a services business.
That matters for US technology companies because it extends the AI investment thesis beyond a handful of frontier labs. If evaluation and training data are treated as durable, defensible businesses, then the number of investable AI companies widens, and so does the set of firms that enterprise buyers must negotiate with.
The unicorn board is being rebuilt younger
The Crunchbase Unicorn Board topped 1,900 private companies in October, a new record, with 34 companies joining and adding more than $96 billion in value and $14.8 billion in total funding to the board, as Crunchbase News reported. The detail that matters most is age: half the new entrants are under four years old. The board also set a US record, with more than 1,000 private unicorn companies.
A young cohort of unicorns is a different kind of animal from the 2021 vintage. Companies that reach a billion-dollar valuation in under four years have typically raised against a narrow, fast-moving thesis rather than a long track record. The record count is therefore a statement about where capital expects growth to come from - and it is consistent with the AI concentration visible in the Mercor round.
The top of the market is setting the terms
Anthropic is seeking to secure up to $88.5 billion in its planned IPO and a concurrent private placement, submitting an updated S-1 with new details on share price range and offering size, as Crunchbase News reported. If completed at that scale, it would be the largest-ever fundraise tied to a public market debut.
This is the anchor for the other two stories. A raise of that size, tied to a public listing, gives the AI sector a public benchmark where previously there was mostly private marking. It also pulls capital toward the top of the market. When the largest private AI companies can absorb tens of billions in a single transaction, the opportunity cost for institutional investors is real: every dollar committed there is a dollar not committed to the next Mercor or the next four-year-old unicorn.
