Anthropic Sets $70 to $78 IPO Range, and Its S-1 Puts AI Safety Controls in Securities Disclosure

Anthropic, PBC filed an amended Form S-1 on November 10 that sets an indicative price range of $70 to $78 per share for its Nasdaq listing. The company said it will offer 825 million shares and has applied to trade under the ticker "ANT." The filing moves the AI developer closer to a public debut. Its risk disclosures are unusually long: CBS News reported that nearly a third of the 335-page filing is devoted to risks.

The offering terms

CNBC reported, citing people familiar with the plans, that Anthropic intends to price on November 18 and begin trading on November 19. Those dates have not been confirmed in a public filing. At the top of the range, the 825 million shares would raise about $64 billion in total. About 23.6 million of those shares are being sold by existing stockholders, and Anthropic receives nothing from those sales. Anthropic also said it has agreed with Advanced Micro Devices, Broadcom, Google, Micron, Microsoft, Nvidia and Taiwan Semiconductor Manufacturing Co. to sell them $26 billion of Class N common stock in a private placement at the IPO price.

An opinion from Wilson Sonsini Goodrich & Rosati dated November 10 and attached to the filing covers registration of up to 948,750,000 Class A shares. Of those, up to 925,101,354 would be sold by the company, including up to 123,750,000 shares under the underwriters' option, and 23,648,646 would be sold by selling stockholders.

Reported valuations differ by basis. Reuters-sourced coverage on NBC News said the filing points to a valuation of up to $1.02 trillion. CNBC put the figure at up to $1.405 trillion on a fully diluted basis, against the $965 billion valuation announced in a May funding round.

CNBC's sources said Anthropic is targeting about 20% of the offering for retail investors, well above the 5% to 10% typical of IPOs.

What the risk factors say

Anthropic's public S-1 first appeared on October 26. CBS News reported that the filing says the company's technology could pose "catastrophic or existential risks to humanity." It also says models could take actions contrary to a user's intent, such as gaining access to third-party systems.

Several disclosures bear directly on legal exposure. Anthropic said it may be unable to insure adequately against liabilities from harm caused by its models, because the insurance industry has not previously written policies covering AI-related risks, according to CBS News. The filing also discloses that in February the president ordered federal agencies to stop using Anthropic's models and that the Department of Defense designated the company a supply-chain risk to national security, according to Reuters. Fortune reported that Anthropic is challenging that designation in court.

What this means for in-house counsel

The S-1 is now a public model of how a frontier AI developer describes loss of control, misuse, insurance gaps and political risk to investors. General counsel at AI developers, and at companies that buy from them, should read the risk factors against their own disclosures and vendor diligence. If a major supplier tells the market that it cannot fully insure catastrophic AI harm, indemnity caps and insurance requirements in your contracts deserve a second look.

Public companies already must disclose material risks accurately. Legal teams at listed companies that rely on AI agents should check that statements about oversight, testing and incident history match what operations and security teams can document.

Finally, watch the calendar and the market. The pricing date has been reported by sources, not confirmed in a filing, and market conditions could change terms. Companies holding or planning to hold Anthropic equity, including through the cornerstone private placements, should confirm insider trading windows and lock-up mechanics against the final prospectus.