Anthropic Warns of Risks of Autonomous AI Development
In a 93-page S-1 form, the company listed hundreds of risks, including the autonomous acceleration of AI development beyond the company's ability to evaluate and safeguard new models.
The U.S. Securities and Exchange Commission published Anthropic's S-1 form ahead of the anticipated initial public offering. In the document, the company combines a forecast of rapid revenue growth with warnings about higher expenses and other business threats. Anthropic also states that AI could transform the global economy more profoundly than industrialization, electricity, semiconductors, and the internet.
In the risks section, the company addresses competition, costs, customer acquisition, and potential regulation, as well as threats associated with future systems. The discussion does not focus on Anthropic's current models but on those that could largely automate AI research and development. According to the company's assessment, with high autonomy, such systems could accelerate AI development to the point where humans would be unable to evaluate new models and ensure their safety quickly enough.
