Anthropic Eyes Blockbuster IPO, Warns AI Could Have ‘Catastrophic’ Consequences

Anthropic is moving towards a potentially record-breaking stock market debut, promoting artificial intelligence as a technology that could help tackle diseases such as cancer while acknowledging that its unchecked development could bring “catastrophic” risks.

The company behind Claude, its AI assistant, has submitted a public filing for an initial public offering (IPO). The listing could reportedly raise as much as $100billion and value Anthropic at approximately $2trillion.

If those figures are achieved, the IPO would rank among the largest in stock market history and become a crucial test of investor enthusiasm for the artificial intelligence sector.

However, confidence in the AI boom is being tested as investors examine the vast amounts of debt being taken on to build the infrastructure needed to support the technology.

Morgan Stanley forecasts that companies linked to AI will borrow about £373billion this year to finance data centres and other projects around the world.

Meta, the owner of Facebook, has separately unveiled plans for a record £71billion fundraising effort as it seeks to accelerate its own artificial intelligence expansion.

Anthropic reported losses of more than $80.7billion during the first half of the year, alongside revenues of $16.3billion. Two unnamed customers accounted for 23 per cent of that income.

The figures are likely to intensify concerns about a potentially circular AI economy, in which suppliers and chipmakers provide funding to AI developers, who then spend that money purchasing products and services from the same businesses.

Susannah Streeter, chief investment strategist at investment platform Wealth Club, said: “Anthropic’s IPO is shaping up to be a major test of investor appetite for AI, especially as it’s arriving at a particularly turbulent time for markets.

“The shine is beginning to fade from parts of the AI investment boom as investors take a closer look at whether the eye-watering sums being committed to the technology will produce returns capable of supporting such lofty valuations.”

Long list of AI risks 

Nearly a quarter of Anthropic’s 400-page IPO prospectus was devoted to outlining potential risks facing the business.

Among those concerns was the possibility that AI models could progress “more rapidly than our ability to develop, validate, and implement corresponding safety, alignment, and security measures”.

Anthropic warned that advanced AI systems could pose “catastrophic or existential risks to humanity” without suitable safeguards, supervision and security. The danger would be particularly serious if the systems were misused, copied, turned into weapons or permitted to function with a high degree of independence.

At the same time, the company described AI as “one of the most consequential technologies in human history”, arguing that it could significantly improve people’s quality of life.

According to the prospectus, AI could speed up scientific breakthroughs and potentially help cure diseases that have affected humanity for thousands of years, including cancer, Alzheimer’s, heart disease, rare genetic conditions and major infectious diseases.

Anthropic has yet to say how many shares it plans to offer investors. The listing is expected to take place after the US midterm elections.

No decision has been announced on the eventual share price.

Streeter said: “The goalposts have shifted significantly for public market investors. A huge pool of private capital has enabled AI businesses to expand rapidly and secure extraordinary valuations well before reaching the stock market.

“As a result, investors who were unable to participate in private markets may be buying at prices that already assume a large measure of future success. Even so, with fear of missing out still influencing sentiment, demand is likely to be strong among those hoping to secure a stake in one of the biggest names in AI.”

Although Anthropic plans to sell some of its shares to the public, it will give its seven co-founders majority voting control over the company, under a plan outlined last month.

In its prospectus, Anthropic said: ‘We believe that our seven co-founders, Dario Amodei, Daniela Amodei, Tom Brown, Jack Clark, Jared Kaplan, Sam McCandlish, and Christopher Olah, as a group, are distinctly equipped to be stewards of our mission.’

Michael Field, chief european equity strategist at Morningstar, said: ‘The governance structure appears designed to limit the ability of shareholders to influence board composition or facilitate a takeover, which could prove a sticking point for some investors.’