Anthropic’s planned Nasdaq listing would turn the maker of Claude into a defining public-market test for artificial intelligence: a company seeking an extraordinary valuation while warning that the technology it develops could create profound harm.

The company’s filing points to striking commercial momentum. Revenue reached $16.3bn in the first half of 2026, though it also recorded an $80.7bn net loss, according to one account of the prospectus.¹ Supporters of the listing see that combination as familiar to a capital-intensive technology business whose spending is aimed at securing computing capacity and future growth. Market strategist Matt Orton said the disclosure should give investors clearer visibility into where revenue growth is coming from.²

The more cautious reading is that the float exposes how much must go right. Anthropic’s revenue is expanding quickly, but it remains dependent on infrastructure, major enterprise customers and continued access to chips and data centres. One analysis noted that five customers account for about a quarter of its business, while the cost of leasing data centres had reached $321.4bn so far this year.³

Both assessments rest on Anthropic’s ability to remain at the technological frontier. The company says its enterprise adoption demonstrates Claude’s strength, while also acknowledging that a substantial loss of customers or competitive ground could weaken the business case. Its prospectus describes AI as “one of the most consequential technologies in human history,”⁴ framing the public offering as more than a conventional tech debut.

The sharpest tension lies in the company’s own risk disclosures. Anthropic warns that sufficiently advanced systems, without safeguards and oversight, could pose “catastrophic or existential risks to humanity.”¹ It has also highlighted potential misuse in cyber operations and dangerous weapons development. The planned IPO therefore asks investors to weigh a fast-growing AI business against the costs—and consequences—of building and controlling increasingly capable systems.