Anthropic’s prospectus devotes 69,800 words, about 30% of its 240,000-word filing, to AI-related risks, including catastrophes, job displacement, rogue agents and industry concentration. The company, a Delaware Public Benefit Corporation, mainly acknowledges risks that could affect its shareholders, while providing little detail about consequences for workers, consumers and other third parties. The AI Risks & Remedies Register developed with Oxfam identifies 103 risks, including excessive reliance on algorithms, deskilling, wrongful arrests, discrimination and copyright-related harm. Settlements have already been reached in several cases, including $300,000 in Detroit, $2.275 million in the SafeRent dispute and $1.5 billion by Anthropic to compensate authors and publishers. The author recommends that Anthropic publish a public risk statement to show how much of its business relies on genuine productivity gains rather than costs shifted to third parties.
Source: Read the original article
