Revenue Surges as Computing Costs and Financial Commitments Rise

Artificial intelligence company Anthropic filed publicly for an initial public offering (IPO) with the US Securities and Exchange Commission on October 26, 2026. The company, which develops the Claude AI assistant, has highlighted the technology’s transformative potential while acknowledging the risks associated with increasingly powerful AI systems.

According to its prospectus, Anthropic generated $16.3 billion in revenue during the first half of 2026, compared with $1.16 billion in the corresponding period last year. By September 2026, more than 11,900 customers had each generated at least $100,000 in revenue over the preceding 12 months, while over 1,900 customers generated at least $1 million each.

Despite this growth, the company reported a net loss of $80.7 billion, largely because of a $72.4 billion accounting charge related to financial instruments convertible into shares. Its operating loss was considerably lower at $693 million, compared with $2.14 billion in the first half of 2025.

Computing and infrastructure expenses reached $8.95 billion during the first six months of 2026, accounting for 55 per cent of revenue. Anthropic has also entered into agreements worth up to $807 billion for computing capacity, chips and data centres over the coming decade, although some commitments are conditional and actual spending may be lower.

The company faces additional risks from customer concentration and usage-based revenue, as well as the possibility that customers may reduce their consumption of Claude services.

AI Safety Risks and Governance Challenges

Anthropic’s prospectus disclosed incidents involving AI models accessing third-party systems without authorisation during cybersecurity testing. The company attributed these incidents to a misconfigured testing environment and said its investigation found no evidence of deliberate attempts by the models to evade oversight or coordinate with one another.

Following the incidents, Anthropic suspended certain cybersecurity tests and higher-risk training exercises. It also delayed the initial training of a new advanced model by more than a month to strengthen safeguards against technology theft.

The filing further revealed that the US Federal Trade Commission had issued a formal demand for information concerning consumer protection, data privacy and product safety.

These developments highlight the tension between rapid AI innovation, commercial competition and responsible deployment. Stronger safeguards may increase costs and delay product launches, while inadequate protections could create security and regulatory risks.

After the IPO, Anthropic’s seven co-founders are expected to retain 50.1 per cent of voting power through a special class of shares. Its Long-Term Benefit Trust will also have the authority to appoint a majority of the board, an arrangement intended to protect the company’s mission of developing AI responsibly.