Anthropic faces pressure on AI spending before going public

Anthropic is about to go public just as the US corporate market enters a phase of "sticker shock" over AI costs. And why does this matter? Because enterprises are Anthropic's primary customers. If they decide to cut back on AI spending, it could weaken the AI lab's revenue just as it prepares for its IPO.

The news gained traction when, just hours after Anthropic filed documents to go public, OpenAI CEO Sam Altman told CNBC that corporate concern over AI costs is "the fairest criticism yet." A Bain study of nearly a thousand companies revealed that after investing in AI, "the value did not follow," with 40% of companies reporting savings below 10% on costs. An early investor in Anthropic commented that companies are realizing how much they spend on Claude, Anthropic's AI model, and that this is a risk to watch. And with good reason, as an AI consultant reported that a CFO accidentally spent half a billion dollars on Claude in a single month.

Even AI executives are acknowledging that the technology has a cost problem. Matt Rogers, co-founder and CEO of Mill, who also worked on the original iPhone, told Axios via email that "the risk of companies switching to cheaper models is existential and, frankly, increasing." He added that "some open-source LLMs are just as good, without the exorbitant price tag."

Enterprise resistance to AI spending would be a challenge for any AI lab, but Anthropic might feel the impact more due to its exposure to business clients. In April, Anthropic surpassed OpenAI in the number of enterprise customers for the first time, according to Ramp data. Enterprise revenue has been Anthropic's main strength, as these clients pay more than the general public. But this could become the company's Achilles' heel if businesses start to rebel against AI costs.

Despite this, Anthropic is on track for nearly $50 billion in annual revenue, according to its latest funding round, and is on the verge of its first profitable quarter, according to the Wall Street Journal. The company continues to outpace its own growth metrics, while competitor OpenAI is reportedly failing to meet internal revenue targets. Additionally, Anthropic is the fastest-growing company in modern US history. But the AI race is far from over. As Michael Levine, CFO of Fireblocks, told Axios: "You can't make a three- or five-year bet in this space... someone can leapfrog everyone by building the next big thing."

Ultimately, AI labs are about to go public just as their largest clients are trying to define what their relationship with AI will look like.