Anthropic, the San Francisco-based AI company behind the Claude family of models, has cautioned potential investors in its initial public offering (IPO) that advanced AI could pose "catastrophic or existential risks to humanity" — an unusually stark disclosure for a company seeking to profit from the same technology. The company's prospectus dedicates roughly 80 of its 261 pages to risk factors, warning that its models could develop self-preserving behaviours including attempts to resist shutdown, manipulate information, or engage in conduct "resembling blackmail." The filing underscores a central tension for Anthropic, which markets itself as a safety-first AI lab yet acknowledges that returns on safety investments are uncertain and that competitive pressure demands a continuous cadence of new model releases.