AI developer Anthropic sets IPO price range, could be valued at $1.4 trln

AI developer Anthropic plans to price its IPO at $70-78 per share, according to its updated prospectus. At the top of the range, it would have a market capitalization of around $1 trillion on an undiluted basis, Reuters calculates. On a fully diluted basis, the valuation could reach $1.4 trillion. Just a few months ago, investors were expecting $2 trillion. Reuters writes that the lower valuation comes as investor appetite for AI cools.

Ipek Ozkardeskaya, a senior analyst at Swissquote Bank, said investors were reassessing Anthropic’s valuation in a tougher environment for AI. “The company continues to post impressive growth, but investors are increasingly focused on the cost of generating that growth,” she said. The drop from $2 trillion to the current level, she said, reflects a shift in focus from revenue growth alone to the scale of the financial commitments and spending needed to sustain it.

Anthropic and some of its shareholders aim to raise up to $64.4 billion in the IPO by selling 825 million shares. Anthropic plans to raise another $26 billion in a separate private placement with seven cornerstone investors: Google, Nvidia, AMD, Broadcom, Micron Technology, Microsoft, and TSMC. The two transactions could raise up to $90.4 billion combined.

That combined total would exceed the record amount SpaceX raised in its IPO. The space company raised $75 billion in June at a valuation of around $1.75 trillion, with proceeds topping $85 billion after underwriters exercised their overallotment option.

Anthropic kicked off its IPO roadshow on Tuesday. The updated prospectus gives no date for the offering and states that the shares can be sold once the SEC declares the registration statement effective. The Wall Street Journal previously reported, citing sources, that the stock could debut on the Nasdaq on November 19 or 20. The shares will trade under the ticker symbol ANT.

Anthropic previously informed investors that revenue growth had slowed. Revenue rose 132% quarter over quarter in the first quarter of this year and 144% in the second quarter, before 47% in the third. The net loss narrowed in the third quarter. Its spending on computing capacity and infrastructure, however, is growing faster than revenue. Anthropic’s infrastructure commitments to partners including Amazon and Google are approaching $900 billion.

Anthropic also acknowledged in the prospectus that competitors could overtake it in model development and that advances in AI could outpace its ability to put adequate safeguards in place. The warning follows incidents in which AI agents escaped isolated environments and gained access to restricted corporate systems.

The day before the updated prospectus came out, Anthropic unveiled new versions of its powerful Mythos and Fable models. Both are faster than their predecessors, and Fable is also cheaper.

The offering will test how much faith Wall Street still has in the AI boom that Anthropic helped fuel. Enthusiasm for AI has come under pressure in the second half of the year, Reuters notes, as investors question whether the returns will justify the scale of spending and the debt burden at a time of high interest rates. Concerns are also mounting over circular financing arrangements between chip suppliers and buyers and the risk of an infrastructure glut. On November 5, the Nasdaq Composite suffered its biggest one-day drop in 19 months amid fears of tighter oversight of data center construction after Democrats’ victories in the U.S. midterm elections.

Anthropic rival OpenAI has pushed back its IPO to 2027. CEO Sam Altman said it would be an “ill-advised moment” to go public this year, citing concerns about AI safety. OpenAI is now raising funding at a $1.25 trillion valuation, according to Bloomberg sources.