Anthropic says AI research could outpace model safeguards

Revenue rose 1,300%; IPO could hit market as early as Nov. 16

Anthropic said in the IPO filing it made public Monday, Oct. 26, with the U.S. Securities and Exchange Commission that it has observed its models contributing to the pace of its own research. The company said in the document that risks from developing AI models without sufficient safeguards would increase as AI systems substantially automate AI research and development if, for example, those systems accelerated AI progress beyond Anthropic’s ability to evaluate and safeguard new models; progress beyond that capacity could adversely affect Anthropic’s business, financial condition and results of operations.

The filing put that warning in a single statement: “We have observed that our models are contributing to the pace of our own research and these risks will increase as AI systems substantially automate AI research and development if, for example, they accelerate AI progress beyond our ability to evaluate and safeguard new models, which could adversely affect our business, financial condition, and results of operations.”

Anthropic paired the research-automation disclosure with its broader concern that developing AI models without sufficient safeguards would both negatively affect its business prospects and “pose catastrophic or existential risks to humanity.” The filing also said “Our development of highly advanced models, platforms, and applications and expansion of use cases could further increase the risk that our models cause harm,” and it identified unfavorable media coverage and public sentiment toward the technology as a downside to its business.

The financial disclosures in the same filing stood apart from those risk factors. Anthropic said it has posted net losses every year since it was founded in 2021, and that in the first half of 2026 it lost $80.7 billion and “may not achieve or sustain profitability in the future.” Against those losses, the company reported revenue of $16.3 billion for the first half of 2026, up 1,300% from the first half of 2025.

Anthropic said 250 of its 14,000 enterprise customers have contributed $10 million or more in annual revenue.

The company’s IPO is looking at a potential valuation of up to $2 trillion when it reaches the market as early as Nov. 16. Anthropic said its stock-market ticker will be ANT on the Nasdaq.

The filing also carried a statement from Anthropic’s seven founders, including CEO Dario Amodei, who wrote in the SEC filing: “We believe that A.I. is one of the most consequential technologies in human history.”

Amodei separately addressed concerns about the fast development and adoption of artificial intelligence in an essay last month, published after former Anthropic researcher Jacob Coxon posted on social media that AI could “kill us all by the end of the decade.”