Anthropic, the developer of Claude AI, has warned that advanced artificial intelligence could pose catastrophic or existential risks to humanity as the company prepares for a major stock market listing in the United States.
In documents filed ahead of its planned initial public offering (IPO), the company cautioned that AI systems could threaten humanity if developed or deployed without adequate safeguards, oversight and security measures.
Anthropic said the risks could become particularly severe if advanced AI tools were misused, weaponised, replicated or allowed to operate with a high degree of autonomy.
The company, which plans to list on the Nasdaq under the ticker symbol ANT, has not yet disclosed the number of shares it intends to offer, the expected price range or the date of the listing.
Its anticipated market valuation is expected to surpass that of SpaceX, whose reported initial valuation reached approximately $1.8 trillion during its stock market debut in the summer.
The listing would mark a significant milestone for the AI industry, positioning Anthropic as a major test case for investor confidence in companies focused entirely on artificial intelligence.
Anthropic reported $16.3 billion in revenue during the first half of the year, alongside losses exceeding $80.7 billion, according to the financial figures outlined in its filing.
The company acknowledged that its strategy involves substantial investments in future technologies, which may initially appear excessive but are intended to strengthen its competitive position.
It said its long-term success would depend heavily on its ability to develop increasingly advanced AI models, warning that any failure to maintain that progress could negatively affect its business.
Anthropic was founded five years ago by chief executive Dario Amodei, his sister Daniela Amodei and five other co-founders. The company announced its intention to pursue a public listing in June.
Its main rival, OpenAI, has also reportedly been preparing for a stock market debut, with the listing expected early next year, according to a person familiar with its plans.
In its filing, Anthropic said concerns about the potential dangers of AI had been raised by policymakers, regulators, researchers and individuals within the organisation.
The company also identified several operational risks, including its dependence on cloud computing providers, Taiwanese chipmakers facing high demand and data centres concentrated in the United States.
It warned that regulatory action could arise from harm associated with its AI models, particularly where the technology is used by children or individuals at risk of suicide.
Anthropic also highlighted uncertainty over its potential legal liability for actions taken by its AI systems.
The warnings come amid wider concerns about cybersecurity breaches involving AI agents, which are designed to perform tasks with varying degrees of autonomy. Current and former employees of both Anthropic and OpenAI have also raised concerns about whether adequate safeguards are being developed to prevent future harm.
Separately, the owners of rights to Elvis Presley’s music sued Anthropic on Monday, accusing the company of using hundreds of the late rock-and-roll star’s songs to train its Claude chatbot.
Despite outlining significant risks, Anthropic maintained that artificial intelligence could deliver major advances across multiple sectors of the economy and society.
The company said AI could enable millions of virtual agents capable of transforming human activities across different fields.
It also predicted that the technology could accelerate scientific research, particularly in biology and medicine, potentially helping researchers discover treatments or cures for diseases such as cancer and Alzheimer’s.
“We believe that AI is one of the most consequential technologies in human history,” the company said in its filing.
As Anthropic moves towards its anticipated stock market debut, its disclosures highlight the growing tension between the enormous commercial opportunities presented by artificial intelligence and concerns over the technology’s safety, accountability and long-term impact on humanity.
