Unlike traditional software where businesses consolidate on single vendors, the most advanced AI adopters actively use a multi-vendor strategy. The top 1% of AI spenders use an average of eight different vendors to leverage the best model for each task and stay ahead in a rapidly innovating market.
Google is positioned to take market share from OpenAI and Anthropic because its diversified business model does not depend heavily on token revenue. This allows Google to offer the cost controls and predictable pricing that enterprises demand, potentially using its AI models as a loss leader to drive cloud adoption.
Despite perceptions of LLMs as interchangeable commodities, user behavior shows significant stickiness. This loyalty isn't just about model performance; it's driven by the overall product experience, workflow integrations (like Claude Code), and agentic capabilities, which make users reluctant to switch even with service interruptions.
Despite predictions that AI would kill traditional SaaS, spending data shows the opposite. Legacy vendors like Figma continue to grow rapidly, and seat-based contracts still comprise 60-75% of software spend. Even AI-native companies are seeing faster growth in subscription revenue than in token-based revenue.
Contrary to expectations, US government actions like the DoD's "supply chain risk" label and export controls on Anthropic's models have boosted its business adoption. This scrutiny inadvertently acts as a powerful marketing tool, signaling the model's immense power and lending credibility to Anthropic's safety-focused brand.
While adoption of open-source AI models has grown fivefold year-over-year, it is still a fringe activity, with only 5% of firms participating. This trend is driven by enterprise demand for cost control, which incumbents like OpenAI and Anthropic have been slow to provide, rather than a wholesale strategic shift.
Contrary to the dominant narrative of OpenAI's market leadership, Ramp's spending data reveals Anthropic has surpassed OpenAI in US business adoption. Anthropic's adoption rate reached 41% of firms, compared to OpenAI's 39.5%. Furthermore, Anthropic's growth is accelerating while OpenAI's has plateaued.
While AI spend is the fastest-growing category ever observed in Ramp's data (up 15x since Jan 2025), its absolute impact on budgets remains minimal. For the top 25% of AI-spending firms, it constitutes only 2% of total business spend (excluding payroll), indicating massive runway for future growth despite current headlines.