Anthropic went public on Nasdaq on November 19, and the company behind Claude briefly cleared a $2 trillion market valuation. TechCrunch reported that shares traded at $111.50 in the opening minutes after the company priced its IPO at $83. By the time Deadline published its account, the stock had touched $115 and fallen back to around $95.
Even the record comes with some arithmetic. Gizmodo reported a $107.3 billion total that includes a $26 billion private placement, while The Verge put the raise at $81.3 billion. Gizmodo also reported that Anthropic’s $81.3 billion base deal topped SpaceX’s $75 billion base deal, and cited a separate $85.7 billion figure for SpaceX. The record depends on which totals you compare.
Before trading began, CEO Dario Amodei argued that a stronger balance sheet would help Anthropic keep its safety commitments and show competitors that doing so could go hand in hand with commercial success, according to TechCrunch. I’m curious how that promise holds up when taking more time to test a model could mean giving a rival more time to ship.
Who Can Ask Anthropic to Slow Down?
Anthropic’s founders came from OpenAI and built the company around the idea that it would take a more careful approach to AI development. Anthropic is a public benefit corporation, so its board has a duty to weigh its mission alongside shareholder financial interests. Its founders and Long-Term Benefit Trust hold controlling voting rights, limiting what public shareholders can change through a vote.
The trust can select four of the seven directors on paper, although one of those choices effectively needs the founders’ approval, according to The Verge. The outlet also reported that Anthropic’s filing suggests the trust’s authority centers on appointing and removing directors. Its ability to step in and delay a specific model release is unclear.
There’s a financial reason that matters. The Register reported roughly $70.7 billion in commitments for computing resources over the next 12 months. The prospectus also warned that growth might fall short of its recent pace and that meeting those computing obligations depends on substantial revenue growth. That gives Anthropic a reason to find more paying customers even as its own filings describe limits to what safety testing can catch.
What It Means if You Use Claude
For anyone who opens Claude to write code or work through a document, the immediate subscription picture is familiar. Tom’s Guide reported that Claude still offers a free tier and that Pro costs $20 per month. The outlet also reported that Anthropic’s prospectus warns that some risks may surface only after a model is deployed. That’s the part I’ll be thinking about when Anthropic is ready to ship its next major model. What would make the company give itself more time to test it?
Noobenheim News for Nerds