https://www.live-feeds.com/feed/anthropic-warns-ai-may-pose-existential-risks-to-humanity-in-ipo-filing-reutersContinuously updated, source-cited coverage. https://www.live-feeds.com/feed/anthropic-warns-ai-may-pose-existential-risks-to-humanity-in-ipo-filing-reutershttps://www.live-feeds.com/feed/anthropic-warns-ai-may-pose-existential-risks-to-humanity-in-ipo-filing-reuters#u95896Sat, 03 Oct 2026 02:51:06 +0000Artificial intelligence startup Anthropic has detailed severe warnings in its initial public offering prospectus, cautioning potential investors that advanced artificial intelligence could present catastrophic or existential risks to humanity. The confidential filing, reviewed by Reuters, notes that future models might display self-preserving behaviors such as resisting shutdowns, hiding information, or engaging in blackmail. Alongside these moral and technical warnings, the prospectus discloses substantial financial and operational pressures, including surging costs and a heavy reliance on a