Enterprise AI budgets are buckling under the weight of massive token consumption, driven by the adoption of agentic AI tools and advanced models. By mid-2026, many companies report being three times over their token budgets, forcing technology leaders to rethink how they govern artificial intelligence spending.
The agentic AI cost multiplier
J.R. Storment, executive director of the FinOps Foundation, highlighted this strain during the FinOps X 2026 conference. He noted that agentic AI introduces compounding costs through automated reasoning loops and retries.
"Adding reasoning and loops and retries and corrections, and all this insanity where a nonhuman - that does not give up, does not rest, and does not tire - keeps hitting the LLMs, keeps retrying, keeps failing," Storment said. "This created a lot of loops and an extravagant amount of token usage."
The financial impact is already visible. In April, Uber CTO Praveen Neppalli Naga told The Information that employee use of Anthropic's Claude Code exhausted the company's AI budget just four months into the year. Uber subsequently capped employee access to these tools in June to rein in expenses. Goldman Sachs researchers project global token usage will multiply 24 times by 2030, reaching 120 quadrillion tokens per month.
Balancing innovation and token economics
Shutterstock provides a clear example of how usage dictates cost. The company onboarded multiple generative AI tools to help customers create content, but quickly found that output tokens drive 75% of its total consumption.
"Understanding your AI costs is no longer optional, it is foundational to business strategy," said Courtney Totten, Shutterstock's CTO and CISO. Totten explained that while less advanced models can handle basic prompts, complex queries require larger, more expensive models.
When the company jumped from 50 to 750 ChatGPT user licenses in a single month, leadership lacked the visibility needed to manage the financial impact. To fix this, Shutterstock's FinOps team began tracking AI costs alongside traditional cloud expenses. This centralized oversight helped the company identify $250,000 in unused vendor commitments that would have otherwise gone to waste.
The shift to strategic FinOps
Traditional cloud cost management is no longer sufficient for modern AI workloads. Pooja Kumar, vice president of cloud strategy and transformation at Prudential Financial, argues that FinOps teams must pivot from building dashboards to owning token economics.
"The questions that I'm getting and executives are asking these days are not about how much I'm spending on cloud, AI or other things, that's not the question anymore," Kumar said. "It's, 'what does one business outcome cost end-to-end and what is the cost of doing it responsibly?'"
This shift requires technology leaders to partner closely with finance. Executives need clarity on which AI providers the company uses, what models are accessible, and where vendor commitments exist to enable accurate cost forecasting. For organizations looking to align these financial controls with broader corporate goals, AI for CFOs training provides frameworks for managing these emerging technology expenses.
Why this matters for executives and strategy
Leaders can no longer treat AI spending as a separate, unmanaged variable. The transition from static cloud costs to dynamic, volume-based token pricing requires new governance models.
Executives must mandate top-down oversight of AI procurement and usage. By routing all artificial intelligence expenses through a centralized FinOps function, companies can prevent budget overruns and ensure that every AI tool delivers measurable business value by eliminating wasted vendor commitments.
Your membership also unlocks:
