Anthropic, an artificial intelligence company founded in 2021 by former OpenAI researchers including CEO Dario Amodei, submitted a detailed S-1 filing as it prepares for a potential initial public offering on the Nasdaq exchange under the ticker ANT. The lengthy filing outlines both the company’s rapid growth and the significant risks associated with advanced AI development.
The San Francisco-based firm reported a striking increase in revenue, generating $16.3 billion in the first half of 2026 compared with just $1.2 billion over the same period in 2025, and a mere $4.1 million in 2023. Despite this growth, Anthropic recorded substantial losses, including an $80.7 billion net loss in the first six months of 2026. Most of this figure stems from a one-time accounting charge related to convertible notes, a type of debt convertible into equity. Operating losses, considered a clearer indicator of ongoing business performance, improved significantly to $693 million in the first half of 2026, down from a $2.2 billion loss in the prior year period.
The company holds considerable financial commitments in computing resources, with fixed or minimum spending obligations totaling $265.6 billion as of June 2026, and an additional $245 billion secured since then through new agreements. These contracts include $44.9 billion with Microsoft, $114.5 billion with Google, $110 billion with Amazon, and a $120.5 billion arrangement with SpaceXAI, which is cancellable with 90 days’ notice starting early next year. Anthropic is also working to reduce reliance on major cloud service providers by developing its own data center infrastructure, highlighted by a recent lease agreement with EverWind Fuels for a $10 billion project in Nova Scotia.
Anthropic’s revenue base remains somewhat concentrated, with its top five customers accounting for about 25 percent of sales in the first nine months of 2026, though no single customer comprised more than 10 percent of revenue in that period. Many of these top customers do not have long-term contracts, posing potential risks to revenue stability.
The filing’s risk factors section, spanning over 90 pages, addresses a wide range of challenges, from typical business risks to profound concerns about AI safety and ethics. The company warns of the potential for “catastrophic AI-related incidents” that could overwhelm even the insurance industry’s capacity to provide coverage. As AI models become increasingly sophisticated and autonomous, Anthropic acknowledges the possibility of misuse by malicious actors, including the development of chemical and biological weapons, and unintended harmful actions by the AI itself.
Anthropic also foresees potential regulatory and societal challenges tied to debates over AI sentience and moral status, which could lead to new constraints on AI development. The company’s long-term success, it says, depends in part on using its AI systems to create increasingly advanced models, a process CEO Amodei has cautioned may eventually outpace human control.
Governance at Anthropic reflects its status as a public benefit corporation, balancing shareholder interests with a broader mission to ensure AI benefits humanity. The company has established a Long-Term Benefit Trust, chaired by former Federal Reserve chair Ben Bernanke, which will appoint a majority of the board. Meanwhile, Anthropic’s seven founders retain voting control of 50.1 percent, enabling them to dominate shareholder decisions. This governance structure limits influence by common shareholders and could complicate efforts to remove directors or management.
Anthropic’s trajectory underscores the accelerating pace and complexity of the AI sector, positioning it as a key competitor alongside firms like OpenAI, which recently postponed its own public listing amid safety concerns. As Anthropic moves toward public markets, its extensive risk disclosures signal the company’s acknowledgment of the broad, unprecedented challenges involved in wielding transformative AI technology.