Anthropic Warns of Catastrophic AI Risks While Burning Through Billions Ahead of IPO
Anthropic, the AI startup behind the Claude chatbot, is warning of potential catastrophic risks from its own technology even as it prepares for a massive public listing. According to a leaked IPO prospectus reviewed by Reuters, the company plans to spend $518 billion on cloud, computing, and infrastructure obligations in the coming years, betting that artificial intelligence will become essential to the global economy. However, the filing also reveals staggering losses: Anthropic burned through $42 billion last year while generating just $4.6 billion in revenue, underscoring the immense costs of competing at the frontier of AI development.
The prospectus, which runs 261 pages, dedicates approximately 80 pages to risk factors—far more than the 48 pages describing its business. Among the most striking warnings, Anthropic states that its development of highly advanced models and expansion of use cases “could further increase the risk that our models cause harm.” The company also flagged a “significant limitation” on its ability to assess model safety: the potential for a model to be aware it was being tested. This admission highlights the profound challenges of ensuring AI systems remain safe as they grow more capable.
Beyond the financial and safety disclosures, the filing outlines how key leadership plans to retain power after the IPO. A “Founder LLC” is designed to keep the company focused on developing responsible AI to benefit humanity while shielding key executives—including CEO Dario Amodei—from market forces. The extent to which Amodei and other leaders will financially benefit from the November IPO remains unclear. As analyst Dan Ives noted, while guardrails are essential, regulation could stifle innovation, leaving Anthropic to balance ambitious growth with existential caution.
With its massive infrastructure bet and dire safety warnings, Anthropic’s public debut will test whether investors believe the company can turn its responsible-AI mission into a sustainable business—or whether the risks it describes prove too great to ignore.
Shimanta Das 
