Anthropic’s recent feud with Trump’s boss may be helping, sales data suggests

Anthropic has a month.

AI Lab finished in May overtaking OpenAI in market share for enterprise computing for the first time, Ramp recently revealed. It raised $65 billion at a $965 billion valuation (and is developing OpenAI) at the end of May, then entered June with a private IPO filing, reportedly on the strength of its first profitable quarter.

Then on Friday, the Trump administration renewed its fight against the model maker by sending a letter demanding that it block non-Americans, including Anthropic employees, from accessing its high-end models: the limited-edition Mythos 5 and a heavily guarded version of the Mythos released to the public three days ago, called the Fable 5.

This forced Anthropic to pull its latest all-powerful model off the market entirely.

Although the White House invoked an ambiguous export control order in ordering the ban, the exact cause remains unclear. The discussion was that hackers were easily bypassing Fable 5’s security controls, which were intended to prevent access to Mythos’ abilities. That model is so good at finding security flaws in software code that Anthropic itself markets it as dangerous and has limited its public release.

The new drama comes after Anthropic refused to allow the government to use its models to monitor most Americans and private weapons. As a result, in March, the Trump administration declared the company at risk of supply.

That didn’t stop Anthropic from selling to businesses. On the contrary, Ramp data shows. Ironically, this latest conflict with the Trump administration, which also seems to confirm the upheaval over Mythos’ legendary power, may help rather than hurt Anthropic, according to Ramp’s lead economist, Ara Kharazian. Kharazian is the inventor of AI data that destroys business.

“If anything, it will improve them,” Kharazian told TechCrunch. “Anthropic’s best month on record, in terms of acquisitions, was the month the Department of Defense called them a supply chain risk. There’s a lot of aura that comes with having your model specifically called out for being too risky to use.”

Ramp’s data isn’t granular enough to see how much money the company will take financially by bringing the Mythos and Fable 5 to market.

Yet data, from more than 70,000 businesses using its platform, shows that customers are increasingly using Anthropic’s Opus models and that business use has been growing.

For example, Ramp reported that Anthropic’s share of AI subscriptions paid to businesses rose 2.5 percentage points in May to 41%. This compares to OpenAI, which commanded 39.5% of AI subscriptions by its customers, basically down from last month. (OpenAI still leads Anthropic in overall consumer spending, according to new data from Sensor Tower.)

Outside of subscriptions, most of what companies spend money on are API calls to the model, which cover the use of tokens for tasks like writing code. Claude’s Anthropic Code has a solid reputation as a powerful AI coding tool.

Ramp can’t always tell from spend data which models most businesses are using. If it can see the details of the model – about one third of the jobs – businesses are using a lot of different flavors of Claude Opus, especially the latest versions. The Opus is the predecessor to the Mythos and is still available in the open.

In fact, in late May, Anthropic released a new version, Opus 4.8.

The Mythos wasn’t in the market that long, having been released to limited users since April. And Fable 5 was closed after a few days.

Although we cannot predict how this latest drama with the White House will affect Anthropic’s ability to go public as expected (public market investors tend to be wary of companies involved in conflicts with the government), the numbers show that Anthropic’s available models are more popular with businesses than ever.