Anthropic filed for its initial public offering just as major corporations are reckoning with sky-high artificial intelligence bills and disappointing returns on their AI investments. The timing collision puts the AI startup in a precarious position as it prepares to go public.
Within hours of submitting IPO paperwork, OpenAI CEO Sam Altman told CNBC that corporate anxiety over AI expenses represents "the most fair criticism of AI so far." A Bain survey of roughly 1,000 companies reinforced the concern: after deploying AI systems, 40% reported cost savings below 10%, with many finding that promised value never materialized.
The cost reality is hitting hard. One Anthropic investor told Axios that enterprises are suddenly conscious of how much they spend on Claude, the company's flagship AI model, and view it as a material risk. A consultant recounted a case where a CFO client accidentally burned through half a billion dollars on Claude in a single month.
Industry insiders are sounding alarms. Matt Rodgers, co-founder and CEO of Mill and former iPhone engineer, warned Axios that "the risk of enterprises switching to cheaper models is existential and, frankly, escalating." Open source large language models are delivering comparable results without the premium price tag, he noted.
What makes this particularly thorny for Anthropic is the company's dependence on business customers. In April, Anthropic overtook OpenAI in the number of enterprise clients for the first time, according to Ramp data. These corporate deals generate far more revenue than consumer use, making them the engine of Anthropic's growth. But that same advantage becomes a vulnerability if businesses cut spending or migrate to cheaper alternatives.
Anthropic's financials paint a contrasting picture. The startup is tracking toward nearly $50 billion in annual revenue based on its latest fundraising round and posted its first profitable quarter ever, according to the Wall Street Journal. The company has consistently beaten its own growth projections while competitor OpenAI reportedly misses internal revenue targets. Anthropic also holds the title of fastest growing company in modern American history.
Yet uncertainty looms over the entire AI landscape. Michael Levine, CFO of Fireblocks, cautioned Axios that long term strategic bets in AI are risky. "Someone can jump over everybody else by coming up with the next great thing," he said, noting the impossibility of making reliable three or five year projections.
The broader pattern is clear: AI companies are heading to public markets at the exact moment their primary revenue source is questioning the financial case for artificial intelligence. Whether Anthropic can persuade Wall Street and enterprise buyers that its technology justifies the cost will largely determine whether the IPO marks the beginning of a growth story or the start of a reckoning.
Author James Rodriguez: "Anthropic's got the best enterprise numbers in the space, but riding those contracts public while CFOs are doing sticker shock audits is textbook bad timing."
Comments