Anthropic's Washington feud is boosting sales, Ramp data shows
A government order to bar non-Americans from Anthropic's top models has not slowed it. Ramp data shows record enterprise demand, and India has the most to lose.
The News
Anthropic, the AI lab led by Dario Amodei, is having its strongest commercial run yet, and the trigger appears to be its escalating standoff with the Trump administration. Spending data published on Friday by the corporate finance platform Ramp shows that Anthropic's share of business AI subscriptions climbed 2.5 percentage points in May to reach 41%. That puts it ahead of OpenAI, which held 39.5% of subscriptions among the same pool of customers.
The figures draw on more than 70,000 businesses that use Ramp's platform, and they point to heavy corporate reliance on Anthropic's Opus models. The commercial momentum mirrors the company's financing: it raised $65 billion at a $965 billion valuation at the end of May, eclipsing OpenAI on that measure too.
The backdrop is a bruising fight with Washington. The administration recently sent Anthropic a letter demanding it block non-Americans from using its most advanced systems, the limited-release Mythos 5 and Fable 5. Rather than comply selectively, Anthropic pulled its newest flagship model from the market entirely. The dispute follows a March designation that branded the firm a supply-chain risk.
Why It Matters
Ramp's lead economist, Ara Kharazian, argued the political heat is not denting demand. "If anything, it'll probably boost them," he said, noting that Anthropic's best month on record coincided with the Department of Defense labelling it a supply-chain risk.
That dynamic is striking. Conventional wisdom holds that a public feud with a sitting government spooks enterprise buyers who crave stability. Here the opposite is happening. Anthropic's earlier refusals to let its models be used for mass surveillance or autonomous weapons have hardened its reputation as the safety-first option, and risk-averse corporate buyers seem to be rewarding that posture rather than fleeing it.
There is a precedent worth recalling. When OpenAI's governance crisis erupted in November 2023 and Sam Altman was briefly removed, many predicted a customer exodus; instead the drama cemented OpenAI's name recognition and accelerated adoption. Controversy, for category-defining AI firms, has so far behaved more like marketing than liability.
Indian Angle
For Indian technology buyers, the more consequential detail is not the sales chart but the export demand. An order forcing Anthropic to bar non-Americans from its frontier models would land directly on India. The country hosts a vast share of the world's global capability centres, and engineers in Bengaluru, Hyderabad and Pune routinely build on Claude and Opus for the very US enterprises driving these subscription numbers. A citizenship test on model access would fracture those workflows overnight.
It also sharpens the case for homegrown capability. Indian foundation-model efforts such as Sarvam and Krutrim have pitched sovereignty as a selling point, and a Washington-imposed access barrier turns that pitch from patriotic talking point into procurement logic. Indian IT services majors that have standardised on US frontier models may now weigh multi-model strategies and on-shore alternatives more seriously.
Regulators will be watching too. MeitY's continuing work on data and AI governance, alongside the questions SEBI-listed IT firms must field on model dependency, makes foreign export controls a live boardroom risk rather than an abstract geopolitical one. For Indian developers paying in rupees for dollar-priced tokens, any supply shock also threatens cost predictability.
FAQ
What did the Trump administration actually demand?
It sent Anthropic a letter requiring the company to stop non-Americans from accessing its state-of-the-art models, the limited-release Mythos 5 and Fable 5. Anthropic responded by withdrawing its newest flagship model from the market rather than enforcing the restriction.
How big is Anthropic's lead over OpenAI now?
On Ramp's data, Anthropic held 41% of business AI subscriptions in May against OpenAI's 39.5%, a gap of 1.5 percentage points after Anthropic gained 2.5 points in the month.
Why would a government feud help sales?
Ramp economist Ara Kharazian suggests the publicity and Anthropic's safety-led positioning reassure cautious enterprise buyers. Its record month coincided with a Defense Department supply-chain-risk label, implying the controversy raised its profile rather than scaring customers away.
What should Indian firms do about it?
Review dependency on US frontier models, evaluate domestic options such as Sarvam and Krutrim, and build multi-model fallbacks so that any citizenship-based access rule does not disrupt India-based development teams.
This story was reported by TechCrunch. Read the full original coverage at TechCrunch.
Sources & Citations
Frequently Asked Questions
What did the Trump administration actually demand?
It sent Anthropic a letter requiring the company to stop non-Americans from accessing its state-of-the-art models, the limited-release Mythos 5 and Fable 5. Anthropic responded by withdrawing its newest flagship model from the market rather than enforcing the restriction.
How big is Anthropic's lead over OpenAI now?
On Ramp's data, Anthropic held 41% of business AI subscriptions in May against OpenAI's 39.5%, a gap of 1.5 percentage points after Anthropic gained 2.5 points in the month.
Why would a government feud help sales?
Ramp economist Ara Kharazian suggests the publicity and Anthropic's safety-led positioning reassure cautious enterprise buyers. Its record month coincided with a Defense Department supply-chain-risk label, implying the controversy raised its profile rather than scaring customers away.
What should Indian firms do about it?
Review dependency on US frontier models, evaluate domestic options such as Sarvam and Krutrim, and build multi-model fallbacks so that any citizenship-based access rule does not disrupt India-based development teams.