The artificial intelligence sector witnessed a historic shift in market dynamics this quarter as Anthropic, the San Francisco-based AI safety and research laboratory, officially overtook OpenAI in business spending market share. According to the latest AI Index report from Ramp, a leading financial operations platform, Anthropic’s share of enterprise AI subscriptions rose to 41% in May, eclipsing OpenAI’s 39.5%. This milestone arrives during a period of unprecedented financial growth for the company, which recently secured $65 billion in new funding, propelling its valuation to a staggering $965 billion. This valuation not only places Anthropic within striking distance of the trillion-dollar "Big Tech" tier but also positions it ahead of its primary rival, OpenAI, in terms of private market capitalization. The surge in investor confidence follows the company’s first-ever profitable quarter and the subsequent filing of confidential paperwork for an initial public offering (IPO), signaling a definitive maturation of the enterprise AI market.

A Chronology of Rapid Expansion and Regulatory Friction

The trajectory of Anthropic throughout 2026 has been characterized by high-stakes technological breakthroughs and increasing friction with federal regulators. To understand the current climate, one must look at the sequence of events that began in early spring. In March 2026, the Trump administration officially designated Anthropic as a "supply-chain risk." This designation followed a public and contentious refusal by the company to allow its large language models (LLMs) to be integrated into domestic mass surveillance programs or fully autonomous weapons systems. Anthropic’s leadership argued that such applications violated the core tenets of "Constitutional AI," the company’s proprietary framework for ensuring model alignment with human values.

By April, Anthropic began a limited-release rollout of its state-of-the-art model, Mythos 5. Marketed as a significant leap over the previous Opus 4.8 architecture, Mythos 5 demonstrated specialized capabilities in identifying vulnerabilities in software code—a feature that Anthropic itself flagged as potentially dangerous if misused. On May 20, the company announced its first profitable quarter, a feat largely attributed to the massive adoption of its Claude Code suite by Fortune 500 companies. This financial success set the stage for the $65 billion funding round and the IPO filing in late May.

However, the momentum faced a significant hurdle in early June. On June 9, Anthropic released Fable 5, a public-facing version of the Mythos architecture designed with stricter safety guardrails. Within three days of its release, reports emerged that hackers had successfully bypassed these guardrails, accessing the underlying high-level capabilities of Mythos. This security breach prompted an immediate and aggressive response from the White House. On the following Friday, the Trump administration issued a formal directive demanding that Anthropic ban all non-Americans—including its own international employees—from accessing the Mythos and Fable 5 models. Invoking obscure export control directives, the government effectively forced the company to pull its most advanced models from the market, citing national security concerns.

Market Share Dynamics: The Ramp Data Analysis

The data provided by Ramp offers a granular look at how Anthropic managed to capture the lead in the enterprise sector despite—or perhaps because of—its regulatory battles. Ramp’s analysis is derived from the spending patterns of more than 70,000 businesses utilizing its financial platform. While OpenAI has long maintained dominance in the consumer space, as evidenced by recent Sensor Tower data showing superior mobile app engagement, the corporate world is pivoting toward Anthropic’s ecosystem.

The 2.5 percentage point increase in Anthropic’s share of AI subscriptions in May is particularly telling. While OpenAI’s growth remained largely flat, businesses increasingly allocated budgets toward Anthropic’s API calls. These calls are the lifeblood of modern enterprise AI integration, powering everything from automated customer service to complex software engineering tasks. A significant portion of this spending is concentrated on the Claude Opus series, specifically the newly released Opus 4.8.

Industry analysts suggest that Anthropic’s reputation for "safety-first" engineering has paradoxically become its greatest marketing asset. As companies become more concerned about data privacy and the ethical implications of AI, Anthropic’s willingness to challenge government mandates on surveillance has bolstered its image as a trustworthy partner for the private sector. Furthermore, the specialized nature of Claude Code has made it the preferred tool for developers, leading to high-volume API consumption that outweighs the revenue from individual seat subscriptions.

The Mythos Controversy and the "Aura of Danger"

The federal intervention regarding Mythos 5 and Fable 5 centers on the model’s proficiency in cybersecurity. Internal testing and external "red-teaming" revealed that the model could identify "zero-day" vulnerabilities in critical infrastructure software at a rate previously unseen in generative AI. While Anthropic intended for these capabilities to be used for defensive patching, the Trump administration expressed concerns that the model could be weaponized by foreign adversaries if any non-citizen had access to the weights or the interface.

The government’s demand for a "citizens-only" access policy is unprecedented in the software industry and has sparked a debate over the future of international collaboration in AI research. Anthropic’s workforce, like most major Silicon Valley firms, is composed of a global talent pool. A ban on non-American employees accessing their own company’s products creates significant operational and legal hurdles.

Interestingly, the "supply-chain risk" label and the subsequent banning of Mythos have not deterred business adoption. Ara Kharazian, lead economist at Ramp, notes that these events may have actually contributed to the company’s recent success. "Anthropic’s best month on record for business adoption was the month the Department of Defense labeled them a risk," Kharazian stated. He describes this phenomenon as the "aura of danger." When a government identifies a model as too powerful or too dangerous for public use, it serves as a signal to the market that the technology is exceptionally capable. For businesses looking for a competitive edge, the perceived "mythological power" of these restricted models becomes a compelling reason to invest in the brand.

Official Responses and Industry Implications

While Anthropic has remained relatively quiet regarding the specifics of its confidential IPO filing, the company has expressed "deep concern" over the White House’s latest directive. In a brief statement, a company spokesperson emphasized that "restricting access based on nationality undermines the collective effort to build safe and beneficial AI and ignores the reality of a globalized tech industry."

The Trump administration, conversely, has doubled down on its "America First" AI policy. Officials from the Department of Commerce have suggested that further restrictions may be forthcoming for any AI lab that refuses to comply with national security standards regarding surveillance and defense integration. This creates a precarious situation for Anthropic’s upcoming IPO. While the company’s financials are stronger than ever, public-market investors are traditionally risk-averse when it comes to companies in the crosshairs of federal regulators. The outcome of this standoff could set a precedent for how the U.S. government interacts with the burgeoning "AI Super-Unicorns."

Future Outlook: The Road to Trillion-Dollar Status

As Anthropic navigates the fallout of the Mythos ban, its immediate strategy appears to focus on the continued stability and improvement of the Opus 4.8 line. Since Opus remains available and is not currently subject to the same level of restriction as the Mythos architecture, it provides a reliable revenue stream to sustain the company through its IPO process.

The broader implications for the AI industry are significant. Anthropic’s rise demonstrates that the "first-mover advantage" held by OpenAI is not insurmountable, particularly in the enterprise sector where reliability, safety, and ethical alignment are prioritized over consumer-facing features. However, the escalating conflict between private AI labs and the federal government suggests that the next phase of AI development will be as much about geopolitical maneuvering as it is about neural network architecture.

For now, Anthropic stands as a titan of the industry, nearing a trillion-dollar valuation while simultaneously being branded a risk to the state. The coming months will determine whether the company can successfully transition to a public entity while maintaining its commitment to its "Constitutional AI" principles in the face of intense political pressure. Regardless of the outcome, the shift in market share reported by Ramp confirms that the battle for AI supremacy has entered a new, more complex chapter.