Anthropic plots october ipo while washington brands it a national-security risk
Anthropic, the ai darling behind Claude, has quietly told Wall Street it wants to go public next month, setting up a high-stakes duel with OpenAI for the first pure-play generative-ai listing and daring regulators who just labeled the company a threat to the Pentagon’s supply chain.
People close to the talks say Goldman Sachs, JPMorgan and Morgan Stanley are jockeying for lead-left roles on what could top a $60 billion raise—an amount that would dwarf any tech debut since Uber and instantly crown the 2021 start-up as one of the most valuable firms on Earth. The same banks are circling OpenAI’s parallel listing plan, according to the same sources, turning autumn into a cage match for investor cash between the two San Francisco rivals born from the same lab.
Valued at $38 billion in February after MGX steered a $3 billion round, Anthropic has spent the summer retrofitting its cap table for the spotlight. Google, Amazon, Microsoft and Nvidia already hold double-digit stakes and have pledged tens of billions in cloud credits and GPUs—firepower the company says it will need to fulfill a $50 billion promise to build domestic ai data centers.
The pentagon calls it a supply-chain threat
While bankers draft prospectuses, the Department of Defense is drafting briefs. Earlier this year the Pentagon invoked a power normally reserved for Chinese vendors, declaring Anthropic’s models a “substantial risk” to national security and banning military use. The company shot back with a federal suit and last week won an injunction, arguing that a blackout would vaporize revenue it can’t afford to lose pre-IPO.
The clash crystallizes the paradox now facing ai founders: the tighter they hug Washington, the faster the revolving door spits them into courtrooms. Anthropic’s entire brand is built on “constitutional ai” and red-line policies against mass surveillance or weapons automation—positions that won it Fortune 500 clients in finance and health care but painted a target on its back inside the defense establishment.
No comment came from Anthropic, Goldman or JPMorgan on Thursday, and final timing remains fluid. Yet the calendar is shrinking: road-show rehearsals are penciled for the week after Labor Day, with price guidance expected before Halloween. If the markets blink, Anthropic can still tap the same strategic investors who already turned its cap table into a who’s-who of big tech. But the IPO drumbeat signals something louder—an insistence that responsible AI, not just scale, can command public-market premiums.
The wager: investors will value safety moats more than war-room contracts. Either way, October just became the month when Wall Street decides whether ethics or artillery sells at a higher multiple.