What's Happening?
Gartner projects that spending on AI agent software will reach $206.5 billion in 2026, a significant increase from $86.4 billion in 2025, and is expected to climb to $376.3 billion in 2027. This rapid growth in agentic AI, which involves AI systems that plan,
reason, and act across multiple steps, is fundamentally disrupting traditional software licensing models. Procurement teams, accustomed to seat-based contracts, are now grappling with token-based pricing, where costs scale with usage rather than headcount. A token, the basic unit an AI model reads and writes, can lead to highly variable bills depending on how intensely an AI tool is used. Gartner estimates that up to $234 billion of enterprise application software spending is at risk from 'agentic arbitrage' by 2030, representing roughly 20% of enterprise SaaS spend. This shift means that AI is not only changing how software is used but also how it is priced, rendering existing procurement strategies potentially outdated.
Why It's Important?
This shift to token-based pricing for AI agent software has profound implications for U.S. businesses, particularly for their IT budgeting and procurement strategies. The unpredictable nature of token consumption, where costs can multiply rapidly based on AI usage, introduces significant financial risk and complexity. Companies that fail to adapt their procurement practices and cost management frameworks could face unexpected budget overruns, as illustrated by Uber burning through its entire 2026 AI budget in four months. This necessitates a move towards FinOps frameworks, providing real-time visibility into AI spending to avoid post-facto invoice surprises. The disruption to traditional SaaS licensing models also means that software vendors must innovate their pricing structures, and businesses must re-evaluate their software acquisition strategies to align with the new economics of AI. This will impact profitability for both software providers and their enterprise customers, driving a need for greater transparency and negotiation around AI consumption metrics.
What's Next?
Organizations will need to establish robust FinOps frameworks to gain real-time visibility into AI token consumption, tracking which teams, workflows, and models are driving costs. This will enable them to implement strategies like model routing, where lower-cost models are used for less complex tasks, optimizing spend. Procurement teams will be tasked with negotiating new vendor contracts that include spending caps, usage alerts, and requirements for token-level usage data to ensure accountability. They will also need to secure the right to adjust model tiers mid-contract to adapt to evolving usage patterns. The development of agentic workflows, which can prepare bounded work for manager approval, will extend account coverage without proportionally increasing staffing, offering a path to efficiency. However, this also means that procurement must evolve from managing fixed-cost licenses to actively managing dynamic, usage-based consumption, treating token economics as a core competency.
Beyond the Headlines
The transition to token-based AI pricing represents a deeper philosophical shift in how value is perceived and exchanged in the software industry. It moves away from a 'per-user' or 'per-license' model to a 'per-action' or 'per-computation' model, reflecting the intrinsic value generated by AI's processing power. This could lead to a more granular and potentially fairer pricing system, but it also places a greater burden on users to understand and manage their AI consumption. The 'agentic arbitrage' risk highlights the potential for AI agents to perform tasks that were previously handled by human-driven software, raising questions about job displacement and the future of work. Furthermore, the need for FinOps and transparent usage data underscores the growing importance of data governance and accountability in the age of AI, where the 'black box' nature of some AI models can make cost attribution challenging. This evolution will likely spur the development of new tools and expertise in AI cost management and optimization.
