Corporate AI Spending Tips Toward Anthropic for the First Time
Corporate AI spending patterns have quietly crossed an important threshold: Anthropic has edged past OpenAI in business adoption metrics based on real purchasing behavior. Credit card data from Ramp’s expense platform shows Anthropic capturing 34.4% of tracked corporate AI spending, compared with OpenAI’s 32.3%. These figures reflect discrete SaaS-style AI purchases made via business credit cards across more than 50,000 companies, rather than large bundled cloud contracts. That distinction matters for understanding business AI market share: this is where teams are consciously choosing vendors like Claude vs ChatGPT, not simply accepting whatever comes with existing infrastructure. Anthropic’s rise is especially striking given its jump from roughly 9% to the mid-30s in just one year, while OpenAI’s share declined and saw its sharpest single-month drop in February. In this slice of the market, Anthropic enterprise adoption has clearly moved into the lead.
Salesforce’s USD 300 Million Token Commitment as a Market Signal
Nothing underscores Anthropic’s momentum more than Salesforce’s plan to spend USD 300 million (approx. RM1.38 billion) on Anthropic AI tokens this year. Tokens are the fundamental billing unit for Claude models, meaning this is a consumption commitment tied directly to real workloads rather than a marketing partnership. Salesforce reports that AI tools have lifted engineering productivity by more than 30%, and AI now powers 30–50% of its overall workload. Its Agentforce AI business has reached about USD 800 million (approx. RM3.68 billion) in annual recurring revenue, and Anthropic’s Claude sits at the core of many of these workflows. The company is also building an “intermediate layer” to route simpler tasks to smaller, cheaper models while reserving Claude for complex jobs, showing how deeply Salesforce is optimizing around Anthropic’s capabilities and pricing model for sustained enterprise-scale usage.
Why Enterprises Are Choosing Claude vs ChatGPT
Ramp’s data suggests Anthropic’s strategy is resonating with the most demanding customers. The company focused first on highly technical users in sectors such as finance, tech, and professional services, then expanded outward. Among first-time AI buyers making explicit vendor choices, Anthropic reportedly wins about 70% of head-to-head evaluations against OpenAI. That success positions Claude as the preferred option in many enterprise AI evaluations, particularly where safety, controllability, and reliability under complex workloads are front-of-mind. While the data excludes large bundled contracts, it captures a crucial dimension of Anthropic enterprise adoption: teams that test tools hands-on and control their own budgets are increasingly selecting Claude over ChatGPT. This pattern also explains why OpenAI experienced its largest recorded single-month share drop, as technical buyers reassess their stacks and reallocate corporate AI spending toward vendors that align better with their governance and performance requirements.
Salesforce Reshapes Its Workforce Around AI-First Engineering
Salesforce is pairing its Anthropic spending with a dramatic reconfiguration of its workforce. The company has announced it will freeze engineering hiring for the first time in 2025, citing AI-driven productivity gains of more than 30%. Around 15,000 engineers are increasingly shifting into supervisory roles, overseeing AI-generated code rather than writing everything manually. Salesforce’s AI integration runs deep: engineering teams use Anthropic’s Claude, OpenAI Codex, Cursor’s coding tools, and the in-house Agentforce platform, plus a new Headless 360 API layer that gives agents like Claude Code direct access to its enterprise stack. AI agents have already helped reduce support staff from 9,000 to 5,000, even as Salesforce plans to add 1,000–2,000 sales roles to help customers adopt AI offerings. The message is clear: human experts still matter, but future growth hinges on orchestrating AI systems at scale.
What Anthropic’s Lead Means for the Business AI Market Share Landscape
Anthropic’s newfound lead in discrete corporate AI spending does not guarantee permanent dominance, but it does reshape the competitive narrative. In the segment where teams select tools intentionally and pay via credit cards, Claude has overtaken ChatGPT and is gaining momentum. Salesforce’s USD 300 million (approx. RM1.38 billion) token commitment further validates Anthropic as a core enterprise vendor, not a niche alternative. At the same time, the data undercounts vendors accessed primarily through bundled cloud services, so the broader market remains more fragmented than these numbers alone suggest. Still, the inflection is significant: when businesses decide how to allocate new corporate AI spending, Anthropic now often starts from a position of strength. For CIOs and CTOs, this shift is a signal to reassess their own vendor mix and governance models as the Claude vs ChatGPT competition intensifies.