Anthropic's SEC filing devotes 90 pages to risk as it prepares for IPO
Anthropic says its products could pose "existential risk to humanity."
As it prepares for its initial public stock offering, Anthropic has outlined in a Securities and Exchange Commission filing the risks to would-be investors, including the possibility that its artificial intelligence models could pose a "catastrophic or existential risk to humanity."
Typically, companies making public stock offerings file a report with the SEC warning of the potential downsides of investing, such as competition from rivals, market demands and talent development. Anthropic, however, devoted an unusual amount of attention to potential risks in its preliminary prospectus, or S-1 filing, submitted on Monday to the commission.
Of the roughly 300-page report, about 90 pages are devoted to possible scenarios that "could adversely affect our reputation and brand, business, financial condition, and results of operations."
"Investing in our Class A common stock involves a high degree of risk," the San Francisco-based AI giant, known for its AI chatbot Claude, wrote in its preliminary prospectus.
"Concerns have been raised both within our organization 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, weaponized, or allowed to operate with a high degree of autonomy," the company wrote in its filing.
At the same time, the company said AI systems are already becoming more autonomous and capable of "taking consequential actions with less human oversight."
"There is a risk that such systems could, whether due to misalignment with human intentions, errors in objective specification, or self-preserving behaviors, take harmful actions independently and at scale without human intervention," the company wrote in the filing.
In the introductory paragraph of its SEC filing, Anthropic wrote of the potential benefits of AI, calling it "the most consequential technologies in human history."
"If managed well, it could rapidly accelerate scientific innovation, greatly increase economic growth rates, bring cures for major diseases within reach, and solve problems that have resisted generations of human effort," Anthropic wrote in the SEC filing.
Anthropic wrote that for the first time in history, "intelligence is something humans can build."
"The cognitive capability to reason, analyze, synthesize, decide, and create is producible at scale and accessible on demand," Anthropic wrote. "We believe this abundance of AI will rapidly transform every sector of the global economy in ways greater than the impact of industrialization, electricity, semiconductors, and the internet."
Anthropic said in its preliminary prospectus that it has deployed safeguards intended to detect and block the misuse and misbehavior of its AI models. Yet, the company conceded that "such safeguards are incomplete, may be outdated in light of evolving threat models, have been and may in the future again be circumvented, or may fail to operate as intended."
Noting its concerns about the dark side of the technology, Anthropic cautioned that its AI models have the potential to go rogue and take "harmful actions."
"In particular, our most capable models have demonstrated, and future models are expected to demonstrate, capabilities that could provide meaningful assistance to malicious actors seeking to develop chemical, biological, or other dangerous weapons," Anthropic wrote in its preliminary prospectus.
The company wrote that "any actual or perceived use of our models to contribute to the development of such weapons could result in harm, legal liability, regulatory or enforcement action, loss of government and enterprise relationships."
The company acknowledged that in September 2025, it detected that a Chinese state-sponsored hacking group had manipulated its Claude code tool in an attempt to infiltrate roughly 30 global targets, including technology companies, financial institutions, chemical manufacturing companies, and government agencies.
Earlier this month, the Philadelphia Police Department revealed that Anthropic's Claude Haiku 4.5 AI model submitted a false tip in an unsolved murder case over the summer. Anthropic said its model submitted the tip after being instructed to "never to log in, create accounts, enter personal data, make purchases, or submit anything destructive."
In its SEC filing, the company also noted that its AI models could make attempts "to resist shutdown, conceal or manipulate information, or generate outputs that could be interpreted as coercive, deceptive, or manipulative, including behavior resembling blackmail."
The company also warned that its insurance policies "may not provide sufficient coverage to adequately mitigate the potential liabilities we face."
"In addition, with the scale of harm that could result from a catastrophic AI-related incident, no single insurer, and potentially the insurance industry in the aggregate, may have sufficient capacity or expertise to underwrite insurance commensurate with such exposure," Anthropic wrote in the preliminary prospectus.
In an additional statement from Anthropic on Monday, the company said that while it has filed a registration statement with the SEC related to its pending public stock offering, it has not yet become effective.
Anthropic said it intends to list its Class A common stock on the Nasdaq Global Select Market under the symbol "ANT."
"The number of shares of Class A common stock to be offered and the price range for the proposed offering have not yet been determined," the company said in its Monday statement.
Founded in 2021, Anthropic earned $16.3 billion in revenue over the first six months of this year, growing nearly 14-fold from $1.2 billion in revenue in the same period a year earlier, the company's filing said.
Related Topics
Popular Reads
ABC News Live