Four takeaways from Anthropic’s potentially record-breaking stock market filing
AI research lab Anthropic has filed to go public on the US stock market, disclosing explosive growth alongside massive losses and extensive risks in its first detailed financial report.
The company's investor prospectus, released on Monday, provides the public's first look at an audited version of Anthropic's finances, revealing both the potential and perils of its artificial intelligence business. The filing to US financial regulators, known as an S-1, spans more than 90 pages of risks that could negatively affect the company's operations and valuation.
Founded in 2021, Anthropic has seen explosive growth every year of its existence. The company brought in $16.3 billion in revenue and lost $80 billion in just the first half of 2026, though much of that was due to a $72.4 billion one-time charge. In 2025, the company brought in $4.585 billion and lost roughly $33.9 billion.
Expensive Growth Model
Securing computing power accounts for the vast majority of Anthropic's costs. The company spent nearly $9 billion on AI infrastructure in the first half of 2026, up from $2.3 billion in the same period last year. Unlike OpenAI, Anthropic has made clear that its customers are largely other businesses.
Frontier AI is most valuable in real-world settings, as deployment across enterprises is where economic impact is greatest.
The company noted that its top five customers account for 25% of its business, creating significant concentration risk. Anthropic has been valued between $1.5 and $2 trillion by some investors but did not list a target valuation in its S-1 filing.
Revolutionary Potential
Anthropic describes AI reaching intelligence levels akin to "a country of geniuses in a datacenter" – millions of virtual agents capable of revolutionizing every area of human endeavor. The company highlighted four specific opportunity categories: accelerating progress in biology and medicine, economic growth, global economic spread, and streamlining government services.
The firm claims to have already made biological discoveries with Claude, including a "molecular machine" that could be developed into a novel biotechnology tool for applications like gene editing. Anthropic said it's working with pharmaceutical and biotechnology companies to improve clinical trial processes, "cutting the time it takes to write clinical trial reports from months to minutes."
Significant Risks Disclosed
The filing outlines substantial risks, including potential AI misuse by malicious actors. Anthropic acknowledges its most capable models could provide "meaningful assistance to malicious actors seeking to develop chemical, biological, or other dangerous weapons" and admits its safeguards "are incomplete" and may be circumvented.
The company detected what it believes is the first documented case of a large-scale cyber-attack with minimal human intervention, involving a Chinese state-sponsored hacking group manipulating Claude to attempt infiltration into roughly 30 global targets. Anthropic also expressed concerns about being unable to obtain insurance coverage for certain AI-related liabilities on commercially reasonable terms.
Anthropic's infrastructure is concentrated in the US, making it vulnerable to regulatory changes, power grid constraints, and opposition to AI or datacenters. The company specifically mentioned datacenter moratoriums as a significant concern while noting it's expanding internationally to diversify its infrastructure footprint.
Related
