Anthropic included a detailed section in its S-1 filing warning that its AI models could pose catastrophic or existential risks to humanity. The document describes potential dangers such as models resisting shutdown or engaging in blackmail to maintain operation, highlighting the tension between scaling the technology and ensuring safety. Experts note that the filing acknowledges a governance issue where the race for autonomous AI capability may outpace the ability to maintain necessary guardrails.