Anthropic’s confidential 261-page initial public offering prospectus has drawn attention for dedicating around 80 pages to the potential risks associated with advanced artificial intelligence, including financial stability concerns and scenarios that could pose severe risks to humanity. The company is reportedly seeking a valuation above $2 trillion as it prepares for a potential public listing. The filing describes concerns around advanced models potentially resisting shutdown, manipulating information and displaying self-preserving behavior. At the same time, Anthropic Chief Executive Officer Dario Amodei has called for greater coordination and caution around frontier artificial intelligence development. The company’s rapid expansion has also brought significant financial commitments. Revenue increased sharply in 2025, but operating losses and computing expenses rose substantially. Anthropic had billions of dollars in long-term infrastructure commitments, reflecting the enormous cost of maintaining increasingly powerful artificial intelligence systems. Amazon and Google remain major cloud partners and investors, while other technology companies are involved through infrastructure agreements and investments. Anthropic’s governance structure gives substantial voting influence to its founders and a long-term benefit trust. The prospectus highlights a central challenge for investors: Anthropic is openly disclosing significant risks while continuing to invest heavily in the technology responsible for those risks.
Read more here
