Anthropic’s IPO prospectus, which spans more than 90 pages, details numerous risks tied to its AI operations, including potential harm from its models ranging from reputational and financial damage to “catastrophic or existential risks to humanity.” The company warns that the complexity of AI, intense competition, market pressure and stakeholder expectations make it difficult to develop and deploy foundation models safely. Anthropic also faces opposition to data-center expansion in 16 states, as well as insufficient electrical capacity and rising energy costs. Finally, legal scrutiny over child safety, slower-than-expected AI demand and customers choosing cheaper or open-source models could limit its growth, while Meta recently agreed to pay approximately 17 billion dollars to settle child-safety litigation.
Source: Read the original article
