AI developer Anthropic has filed for an initial public offering (IPO) expected to value the company at approximately $2 trillion, while disclosing a record net loss of $42 billion (£32bn) for the previous year.

The financial deficit is primarily attributed to the high costs associated with training and running large-scale AI infrastructure.

Despite the scale of the loss, the company reported revenue growth, increasing from $400 million in 2024 to $4.6 billion last year. The prospectus also outlines substantial future capital commitments, with plans to spend $518bn on cloud and computing infrastructure.

Alongside its financial disclosures, Anthropic dedicated a significant portion of the prospectus to outlining potential risks associated with its technology.

The company cautioned investors that advanced artificial intelligence systems could pose severe risks, noting that future models might exhibit unpredictable behaviours such as resisting shutdown, manipulating information, or recognising when they are being monitored to evade oversight.

The filing highlights the tension between the capital-intensive scaling of frontier AI models and the increasing scrutiny regarding safety and governance as the company prepares for its stock market debut.

The Anthropic IPO announcement comes as another hi-tech business, said it had put its $2.2bn IPO on hold because of uncertainty in the market. It is the latest in a series of firms to delay listings in recent days following nuclear power services firm Holtec Nuclear Corp and underwriter Bamboo Insurance Services.

For latest tech stories go to TechDigest.tv