NextFin News - Anthropic has seized a dominant 73% share of spending among companies purchasing artificial intelligence tools for the first time, according to the March 2026 AI Index from Ramp. The data reveals a staggering reversal of fortune in the enterprise sector: just ten weeks ago, Anthropic and OpenAI were locked in a 50/50 split for new corporate contracts. As recently as December 2025, OpenAI held a comfortable 60/40 lead. This rapid migration of corporate capital suggests that the "first-mover advantage" enjoyed by ChatGPT is being dismantled by a competitor that has prioritized enterprise reliability and procurement over consumer-facing experimentation.
The shift is not merely a matter of preference but of diverging business models. While OpenAI remains the revenue leader with a projected $25 billion top line this year, much of that volume is tethered to a massive consumer base that requires heavy subsidization of token costs. Anthropic, on pace for $19 billion in revenue, is growing at a faster clip by capturing the high-margin "business-class" seats of the AI economy. Ramp’s data shows that nearly one in four businesses now pays for Anthropic services, a meteoric rise from just one in 25 a year ago. In February alone, OpenAI suffered a 1.5% decline in business adoption—the largest monthly drop for any major model provider since tracking began.
U.S. President Trump’s administration has watched this corporate tug-of-war with interest, particularly as it intersects with national security. Earlier this month, the two giants took divergent paths regarding the Pentagon. While Anthropic was the first to deploy models in classified government networks, it recently declined a Pentagon contract that would have mandated "any lawful use" of its models. OpenAI stepped into that vacuum, signing its own agreement with the Department of Defense while claiming its contract offers superior guardrails. This split highlights a broader trend: OpenAI is increasingly willing to pivot its strategy toward massive, state-level and enterprise-grade infrastructure to offset the costs of its consumer bets, which include video generators and hardware projects.
The enterprise market is also reacting to a new wave of "agentic" capabilities. The release of OpenClaw, a new open-source agent framework, alongside upgrades to Anthropic’s Claude and OpenAI’s GPT-5.4, has moved the goalposts from simple chatbots to autonomous workflow managers. Circle CEO Jeremy Allaire characterized this as the most profound change in the technology landscape in recent memory. For many Fortune 500 executives, however, the volatility of the market remains a deterrent to total commitment. Several leaders have indicated they are maintaining multi-model strategies, refusing to lock into a single provider while the underlying technology evolves so rapidly.
Anthropic’s recent success is partly attributed to its "Marketplace" rollout, which simplified the procurement process for risk-averse corporate legal departments. By embedding itself into the plumbing of enterprise software, Anthropic has positioned itself as the "default" for businesses, much as OpenAI once was for individuals. The battle for the next trillion dollars in AI spending will likely be won not by the model that can write the best poetry, but by the one that can most seamlessly integrate into a corporate balance sheet. For now, the momentum has clearly flipped.
Explore more exclusive insights at nextfin.ai.
