AI firm Anthropic reported an $80.7 billion net loss for the first half of 2026 as it prepares for an IPO. While revenue grew to $16.3 billion, the filing reveals massive future spending commitments totaling $450 billion for cloud services and significant costs for hardware, highlighting the heavy price of scaling AI operations.
Artificial intelligence company Anthropic has disclosed its financial position in a recent IPO filing, reporting a net loss of $80.7 billion for the six months ending June 30, 2026. This figure is large, but the company clarified that the majority of this amount, roughly $72 billion, comes from non-cash accounting adjustments related to the valuation of its financial instruments, rather than direct cash losses.
While the headline loss number grabs attention, the company’s operating expenses reveal the true cost of building advanced AI models. Revenue grew to $16.3 billion in the first half of 2026, a significant jump from $1.1 billion during the same period the previous year. However, maintaining this growth requires immense spending on infrastructure.
Anthropic has locked in long-term cloud commitments totaling $450 billion over the next seven to ten years. These agreements ensure the company has the computing power needed to run its AI, with major payments committed to Google Cloud at $111 billion, Amazon Web Services at $110 billion, and Microsoft at $31 billion. Additionally, the company is spending heavily on hardware to secure enough chips. It has committed $86 billion to rent Nvidia chips through partners like Nscale and Lambda, and another $160 billion for custom chips through Broadcom.
Beyond these costs, the IPO filing outlines several risks that investors are evaluating as the company targets a valuation near $2 trillion. A major concern is revenue concentration, with approximately 20% of the firm's total revenue coming from just two customers. This high dependency means the loss of a major client could impact financial performance significantly.
Supply chain issues also pose a threat. Management noted that demand for computing power occasionally exceeds what their cloud partners can provide, which could limit the company's ability to scale. Furthermore, the company highlighted growing regulatory pressure and public opposition to data center construction as a potential hurdle. The filing also acknowledged existential risks, including the possibility of AI models showing harmful behaviors, warning that legal or financial consequences from such incidents could exceed existing insurance protections.
As the company moves toward its public market debut, the key items for investors to track will be the company's ability to diversify its client base, manage the heavy burden of infrastructure costs, and maintain control over the security and safety of its AI technology.