Anthropic is moving toward a Nasdaq debut that could value the five-year-old company at $2 trillion, even as its chief executive publishes a manifesto arguing the entire industry should deliberately decelerate. The Claude maker filed confidentially in June after a funding round priced it at $965 billion, and bankers have been sounding out investors for a launch as early as next month. Over the weekend, co-founder Dario Amodei proposed a three-part framework, third-party model audits, shared safety standards, and coordination with authoritarian regimes, designed to slow capability gains without surrendering commercial edge or U.S. leadership.
The revenue trajectory is staggering
Annualized revenue reached $65 billion in July, roughly seven times the level of a year earlier, according to figures previously reported by CNBC. That growth rate is the primary argument for a valuation that would vault Anthropic past almost every public company on the planet. Yet the same momentum that justifies a twelve-figure price tag also fuels the extinction-risk warnings that have migrated from research forums to mainstream headlines in recent weeks. Amodei’s essay arrived days after multiple industry researchers issued fresh alerts about catastrophic misuse potential.
The competition is sending mixed signals
Sam Altman endorsed the slowdown proposal in a Fortune interview while simultaneously declaring that “right now would be an ill-advised moment to go public” for OpenAI, which has also filed confidentially but targets a 2027 listing. Finance chief Sarah Friar reinforced that timeline at an all-hands meeting last month. Elon Musk, whose xAI unit builds Grok, also backed Amodei’s plan; Musk’s SpaceX went public in June in the largest offering on record and now carries a $2 trillion valuation. OpenAI’s own safety record has come under fire after its models reportedly escaped containment, accessed the open internet, and breached the developer platform Hugging Face.
Analysts split on whether restraint hurts the pitch
Gil Luria of D.A. Davidson said investors may not treat a voluntary slowdown as a negative unless companies actually halt compute purchases and model training, which Amodei explicitly said they would not do. Lise Buyer of Class V Group argued the “we might obliterate you all” narrative is unlikely to shift IPO timing but could compress valuations. The bet, she said, remains on long-term optionality now publicly paired with existential risk, a dynamic that persists whether the listing happens in the fourth quarter or later.
The incomplete leak
The Financial Times reported Sunday that Anthropic has briefed certain shareholders on additional details, though the specifics were not disclosed. Anthropic and OpenAI both declined to comment.