Anthropic, the developer of the artificial intelligence (AI) model "Claude," has begun the process to list on the U.S. stock market.
Anthropic said on the 1st (local time) that it confidentially submitted a draft registration statement to the U.S. Securities and Exchange Commission (SEC) for an initial public offering (IPO). A confidential filing is a process in which corporations first undergo SEC review without disclosing financial information and business details; once the review is completed, they can pursue a listing depending on market conditions.
Anthropic said, "Once the SEC review is completed, we will have the option to decide whether to list," and added, "Whether to proceed with an actual public offering will be determined by market conditions and other factors."
The number of shares to be offered and the offering price have not yet been determined.
Anthropic attracted a new $65 billion investment (about 100 trillion won) last month, giving it a corporate valuation of $965 billion (about 1,460 trillion won). This surpasses OpenAI, which was valued at $852 billion (about 1,300 trillion won) in March.
The two companies, which are aiming to list within the year, are reported to be competing to take the lead in the IPO market. With Cerebras having gone public earlier and Elon Musk's SpaceX preparing for a listing, the move is seen as an attempt to accelerate the IPO timeline to secure investment funds first.
Gil Luria, an analyst at investment bank D.A. Davidson, told Reuters, "OpenAI and Anthropic are competing to list before their capital runs out." He added, "Another reason Anthropic may seek to enter the public market before OpenAI is that it can set financial reporting standards for cutting-edge AI models," and said, "By doing so, it can tilt the standards in its favor."
OpenAI is also preparing related documents, including a prospectus, for submission to the securities authorities, but it is not yet at the stage of an official filing.
Meanwhile, according to AFP, Anthropic will also grant the European Union (EU) access to the "Claude Mythos" model, which has expert-level capabilities for detecting cybersecurity vulnerabilities.
Anthropic recently said it plans to release a Mythos-class AI model to the public within a few weeks after developing security measures.