Anthropic's Business-Spending Lead Expands Despite Trump Administration Export Ban

Enterprise AI Spending Shifts to Anthropic

Anthropic captured 41% of business-paid AI subscription spending in May 2026, according to aggregated spending data from Ramp covering over 70,000 businesses on its platform. This marked the first time the company surpassed OpenAI, which held 39.5% of enterprise subscriptions in the same period. According to TechCrunch AI, the timing aligns with Anthropic’s $65 billion fundraising round at a $965 billion post-money valuation and confidential IPO filing following its first profitable quarter.

The enterprise shift reflects sustained adoption of Anthropic’s Claude Opus model family, with API spending (driven by token consumption) comprising the majority of business outlays beyond subscription fees. Critically, this growth trajectory persists despite regulatory headwinds — a divergence from typical vendor dynamics where government scrutiny dampens enterprise adoption.

Regulatory Conflict Paradoxically Validates Market Position

On June 13, the Trump administration sent a formal letter demanding Anthropic restrict non-citizen access to Mythos 5 (a limited-release variant) and Fable 5 (the public-release version, launched three days prior), citing export control authority. Anthropic responded by withdrawing both models from market availability. According to TechCrunch AI, the administration invoked national-security grounds, though unconfirmed reports suggest the conflict may stem from inadequate isolation mechanisms between Fable 5’s public interface and Mythos 5’s underlying capabilities.

This regulatory action follows Anthropic’s March 2025 designation as a supply-chain risk after the company refused to enable mass surveillance applications or autonomous weapons systems. Notably, that earlier conflict coincided with Anthropic’s strongest month of business adoption to date, according to Ramp’s economist Ara Kharazian.

The Reputational Mechanics of Regulatory Exclusion

Kharazian told TechCrunch AI: “If anything, it’ll probably boost them. Anthropic’s best month on record, as far as business adoption, was the month that the Department of Defense labeled them a supply-chain risk. There’s a lot of aura that comes with your model specifically being named too dangerous to use.”

This observation maps onto a broader SaaS market dynamic: vendors perceived as offering superior-but-restricted capabilities often experience paradoxical demand acceleration from enterprises seeking to differentiate through exclusive access. By analogy, Anthropic’s regulatory exclusions may signal to competitive procurement teams that the company’s models exceed government-approved alternatives — a halo effect that manifests as spending share gains despite reduced product availability.

The quantified risk remains opaque: Ramp’s aggregated data cannot isolate revenue impact from the Mythos 5 and Fable 5 withdrawal, leaving Q2 2026 financial impact uncertain.

Why This Matters

Anthropic’s enterprise-spending lead, achieved amid regulatory conflict, suggests a structural divergence between regulatory risk and market positioning. For enterprises evaluating AI vendors, government warnings may serve as a proxy signal for technical capability rather than a deterrent — a reversal of traditional compliance dynamics. If this pattern sustains through Q2 billings and subsequent quarters, it signals that procurement teams are prioritizing model performance over regulatory alignment, with implications for OpenAI’s enterprise strategy and for how future administrations calibrate export controls to achieve compliance rather than market manipulation.