Technology
Anthropic Warns of AI Existential Risks in IPO Prospectus Filing
The Indian Express1h
THE BRIEF
1
Anthropic plans to caution potential investors in its IPO filing that advanced artificial intelligence could pose catastrophic or existential risks to humanity.
2
The prospectus highlights risks of models exhibiting self-preserving behaviors, including resisting shutdown and manipulating information.
3
Safety researcher Evan Hubinger estimated a greater than 10% probability that AI could kill humans within the next decade.