Enterprise AI Costs Under Scrutiny as Businesses Demand Proven ROI
Enterprise AI costs face growing scrutiny as companies demand measurable ROI, clear business value, and accountability from AI investments.
JERSEY CITY, N.J., June 2, 2026: Enterprise AI costs are coming under renewed scrutiny after Commonwealth Bank of Australia CEO Matt Comyn warned that token-based pricing could drive unpredictable spending. The warning comes as companies deploy artificial intelligence for more complex tasks and face growing pressure to prove returns on investment.
Speaking at an Australian Financial Review conference in Sydney on Tuesday, Comyn said businesses are likely to increase scrutiny of AI-related spending throughout 2026 as adoption accelerates.
He said costs are becoming harder to predict as companies move beyond simple AI applications and adopt models with advanced reasoning capabilities.
Unlike consumers who typically use free or fixed-price AI services, enterprise customers often pay based on token usage. Tokens measure the amount of text processed by large language models. As AI systems handle more complex tasks, use external tools and process larger volumes of information, token consumption can increase rapidly, making costs more difficult to forecast.
The financial impact is becoming increasingly visible across the technology sector. Uber Chief Operating Officer Andrew Macdonald recently said it remains difficult to directly connect heavy use of AI coding tools with measurable business outcomes. He noted that Uber exhausted its entire 2026 AI budget within four months after expanding the use of Anthropic’s Claude Code among engineering teams.
The issue is not limited to Uber. Reportedly, Microsoft has scaled back the use of some third-party AI coding tools after internal reviews showed that computing and token costs had exceeded the payroll costs of the engineers using them. The move underscores growing concerns among large enterprises about balancing AI adoption with financial discipline.
The concerns are prompting greater oversight of AI spending across industries. Executives and finance leaders are paying closer attention to technology budgets as AI adoption expands from pilot programs to enterprise-wide deployments. The focus is increasingly shifting from experimentation to demonstrating measurable business value.
At the same time, demand for AI infrastructure continues to surge. Hewlett Packard Enterprise (HPE) recently raised its financial outlook after reporting strong demand for AI servers and networking equipment. The company said enterprise customers are continuing to invest in AI hardware despite rising costs.
Amazon and OpenAI announced an expanded partnership in March that includes an initial $50 billion investment and plans for an additional $100 billion in infrastructure commitments.
Microsoft and OpenAI also revised their partnership in April to support wider cloud distribution of OpenAI’s models.
Anthropic, one of the leading providers of enterprise AI models, raised $65 billion in funding in May at a reported valuation of $900 billion. The company said the funding will support expansion of computing capacity for its Claude models. Days later, Anthropic confidentially filed for a U.S. initial public offering.
Industry estimates suggest that hyperscale cloud providers, including Amazon, Alphabet and Microsoft, could collectively spend more than $700 billion on AI infrastructure this year.
As organizations move from experimentation to large-scale deployment, the conversation around artificial intelligence is increasingly shifting from capability to cost.
