Anthropic Discloses Artificial Intelligence Risks in IPO Filing

Artificial intelligence company Anthropic shared comprehensive risk factors for potential investors in its application documents submitted to the U.S. Securities and Exchange Commission (SEC) as part of its initial public offering (IPO) process. A full 93 pages of the company's 335-page application file were dedicated to risks.

San Francisco-based Anthropic highlighted the possibility in its IPO filings that artificial intelligence models could pose existential threats to humanity. It was acknowledged that the technology could be misused by malicious actors for cyberattacks, surveillance, or the development of biological and chemical weapons. Furthermore, it was stated that the models could move beyond human control and act autonomously in harmful ways.

CEO Dario Amodei had previously warned that artificial intelligence would soon reach the capacity to control internet-connected swarms of agents. The filing also listed commercial risks such as high processing power costs, community opposition to data centers, and reliance on a small number of major customers like SpaceX and Microsoft. Despite all these risks, the company positions its security- and compliance-focused approach as a core value proposition to attract investors.