Your Money

Never mind the ‘risk to humanity’, Anthropic’s IPO raises more prosaic concerns

Anthropic has plans to raise up to $100bn, part of a trend that could result in too much stock coming to market

Anthropic’s Code W/ Claude event in San Francisco on May 7th. The company Anthropic has plans to raise up to $100bn via an IPO. Photograph: Jason Henry/The New York Times
Anthropic’s Code W/ Claude event in San Francisco on May 7th. The company Anthropic has plans to raise up to $100bn via an IPO. Photograph: Jason Henry/The New York Times
Proinsias O'Mahony's picture
Proinsias O'Mahony
Sat Oct 10 2026 - 06:17

Anthropic’s IPO (initial public offering) prospectus warns its AI model could cause a “catastrophic or existential risk to humanity”.

Some investors may have more prosaic concerns, not least the amount of new stock coming to market.

Anthropic is expected to make its market debut next month, and it’s already clear that building a $2 trillion AI giant requires a very large cheque.

Revenue rose twelvefold last year to almost $4.6 billion but the company still recorded operating losses of more than $8 billion.

READ MORE

More strikingly, it envisages spending $518 billion on cloud computing and infrastructure in coming years. Extraordinary amounts of capital are therefore needed. Saudi Aramco’s $29 billion flotation in 2019 was the previous record, but SpaceX tripled it in June when it raised $86 billion at a $1.77 trillion valuation.

Now Anthropic will probably blow past even that, with plans to raise up to $100 billion at a valuation of about $2 trillion. And there’s more, with OpenAI expected to follow Anthropic with a trillion-dollar listing in early 2027.

The implications extend beyond the companies themselves. In a recent note, investment manager GMO warned that the sheer amount of new equity coming to market could become a catalyst for an AI bubble break.

For almost all of the past 20 years, US companies have bought back more shares than they issued. That is now going into reverse, with hyperscalers such as Alphabet and Oracle issuing shares to fund AI spending, a wave of giant IPOs adding new stock, and roughly $2 trillion of SpaceX shares becoming freely tradeable by June 2027.

GMO estimates that, if history is any guide, the resulting increase in equity supply could leave the market about 20 per cent below where it otherwise would have been over the following 18 months.

Today’s AI-driven market may therefore face a more prosaic threat than disappointing or even dangerous technology: simply having too much stock coming to market.

READ NEXT