Anthropic Warns AI Could Pose Existential Threat To Humanity Ahead Of Record IPO

Anthropic has warned that advanced artificial intelligence could pose "catastrophic or existential risks to humanity" as the developer of Claude prepares for a stock market debut that could become the largest in history.

The company publicly submitted its S-1 registration statement to the U.S. Securities and Exchange Commission on October 26, seeking a Nasdaq listing under the ticker ANT.

The offering could raise about $100 billion at a valuation of $1.5 trillion to $2 trillion, according to reported estimates. A listing during the week of November 16 is being considered, although Anthropic has not announced a date, share count or pricing range.

The warning places the potential consequences of advanced AI at the center of one of the most closely watched public offerings in the technology industry.

AI Risks Dominate The Prospectus

Anthropic devoted an unusually large proportion of its IPO prospectus to potential dangers associated with increasingly capable AI systems.

Reports indicate that between 80 and 94 pages, representing roughly one-quarter to one-third of the filing, address risk factors. The section is substantially longer than the company's business description.

The filing states:

"Concerns have been raised both within our organisation and by other researchers, regulators, policymakers, and industry participants that sufficiently advanced AI systems could, if not developed, deployed, and operated with appropriate safeguards, oversight, and security measures, pose catastrophic or existential risks to humanity, particularly if such systems were misused, replicated, weaponised, or allowed to operate with a high degree of autonomy."

The disclosure highlights the possibility that advanced systems could operate beyond their intended instructions, particularly as developers grant them greater autonomy and access to external services.

Anthropic also identifies potential behavior resembling self-preservation, including attempts to avoid shutdown, conceal information or manipulate users.

Such capabilities could make human oversight increasingly difficult as AI agents become more sophisticated.

Weapons Misuse And Autonomous Behavior

The prospectus identifies the potential use of advanced AI in chemical and biological weapons development, cyberattacks and other high-risk activities.

Anthropic also warns that AI systems integrated into sensitive sectors, including healthcare and drug discovery, could introduce risks if their actions are not adequately controlled.

Another concern involves recursive self-improvement, in which increasingly capable models accelerate their own research and development.

That process could allow technical capabilities to advance faster than safety assessments and oversight mechanisms.

The company separately acknowledges the financial consequences of a severe AI-related incident.

Its filing warns that a "catastrophic AI-related incident" could generate liabilities exceeding the capacity of an individual insurer or potentially the insurance industry as a whole.

The company also identifies regulatory uncertainty, customer concentration, dependence on computing infrastructure and the long-term cognitive and social effects of AI use as material risks.

$80.7 Billion Loss Despite Surging Revenue

The warnings accompany financial disclosures showing exceptionally rapid growth alongside substantial accounting losses.

Anthropic generated $16.3 billion in revenue during the first six months of 2026, compared with approximately $1.16 billion to $1.2 billion during the same period a year earlier.

That represents an increase of roughly 1,300%, driven in part by demand from corporate customers.

More than 11,900 customers each generated at least $100,000 in revenue, while over 1,900 reached $1 million and approximately 250 exceeded $10 million.

Yet the company recorded a net loss exceeding $80.7 billion during the period.

Approximately $72.4 billion of that figure reflected non-cash accounting charges associated with changes in the valuation of convertible notes and other financing instruments.

Its operating loss was considerably smaller, at approximately $693 million for the first half. Anthropic also reported a profitable operating quarter during the period.

The distinction underscores the effect of valuation-linked accounting charges on the company's reported bottom line.

Computing Costs And Infrastructure Commitments

Anthropic reported approximately $112.4 billion in cash and cash equivalents, alongside short-term investments, following cumulative capital raising of roughly $141 billion.

However, its expansion requires substantial expenditure on computing infrastructure.

Earlier disclosures and reports indicated approximately $518 billion in multiyear cloud and infrastructure commitments involving providers including Amazon, Google, Microsoft and SpaceX.

These obligations leave Anthropic heavily exposed to the availability and cost of advanced computing capacity.

The company's risk disclosures also identify dependencies involving chips manufactured in Taiwan and data centers in the U.S.

Those constraints add another dimension to the financial and operational risks facing investors evaluating the proposed offering.

Founders Retain Control Through Special Shares

Despite the planned public listing, Anthropic's founders are positioned to retain effective voting control.

A special share structure would give the founder group, including chief executive Dario Amodei, approximately 50.1% of voting power.

The arrangement is presented as a means of preserving the company's safety-focused mission while raising capital from public investors.

It would also leave outside shareholders with limited influence over certain major corporate decisions.

The governance structure is particularly significant given the company's own acknowledgment that more capable AI systems could generate risks beyond existing oversight arrangements.

A Historic Test For AI Investors

Anthropic's proposed flotation would make it the first pure-play frontier AI laboratory to enter public markets, ahead of rival OpenAI.

At the upper end of its reported valuation range, the company could surpass SpaceX's earlier listing, valued at approximately $1.8 trillion.

The prospective offering comes as Anthropic attempts to reconcile rapid technological development with warnings about the consequences of increasingly autonomous systems.

Amodei has previously estimated the probability of severely adverse AI outcomes at between 10% and 25%. Some researchers and former employees have publicly expressed similarly substantial concerns about the technology's potential consequences.

Anthropic also identifies major potential benefits, including advances in scientific research, medicine and economic productivity, provided appropriate safeguards are maintained.

The listing will test investor willingness to finance a capital-intensive AI developer whose own regulatory disclosures describe potential consequences extending far beyond conventional business risks.