AI firm Anthropic says its systems could pose a 'catastrophic' risk to humanity

Anthropic warned in documents filed ahead of its planned initial public offering that its AI systems could pose severe risks, while also describing potential benefits for scientific research. The supplied coverage says the listing would raise capital for computing infrastructure and allow existing investors to sell shares.

• Anthropic’s filings warned that its increasingly powerful systems could pose catastrophic or existential risks, including through biological or chemical weapons and unpredictable autonomous behavior.· 1 source
• The Independent article says nearly a quarter of the documents were devoted to warnings, including unknown long-term effects of regular AI use on mental health and cognitive development.· 1 source
• The article says Anthropic also argued that AI could accelerate scientific research and potentially help address diseases such as cancer and Alzheimer’s.· 1 source