Enterprise artificial intelligence spending has reached 3.3% of corporate revenue in 2026, forcing banks and large corporations across the United States and globally to overhaul how they budget for, govern, and measure technology deployments. As organizations transition from small-scale pilot programs to enterprise-wide systems, escalating expenses for compute infrastructure, application programming interfaces (APIs), and per-token consumption are outpacing traditional IT budget structures.

A recent series of October 2026 industry analyses—including studies from Boston Consulting Group (BCG), Wharton Human-AI Research, and Javelin Strategy & Research—shows that while corporate AI budgets continue to expand, the era of unmonitored experimentation is ending. Financial institutions and corporate executives are now instituting stricter financial operations controls, centralizing C-suite oversight, and tying future capital allocation to measurable revenue growth rather than simple time savings.

From Flat-Rate Subscriptions to Token Economics

During the initial wave of enterprise AI adoption, costs remained relatively predictable. Employees primarily used consumer-grade or basic business chat interfaces on an ad hoc basis for individual productivity tasks, such as drafting emails or formatting presentations. According to PaymentsJournal, citing a report by Matthew Gaughan, Tech & Infrastructure Analyst at Javelin Strategy & Research, early access was heavily subsidized by AI vendors’ venture funding and sold via low-cost, flat-rate subscriptions.

As banks and enterprises formalized their deployments—building custom fraud mitigation tools, software development pipelines, and automated agentic workflows—the underlying pricing architecture shifted. AI labs moved toward usage-based enterprise pricing models centered on per-token consumption and dedicated engineering support, including forward-deployed engineers who build custom systems alongside internal corporate teams.

Those variable costs are now creating friction across corporate balance sheets. According to Yahoo Finance, which reported on the fourth annual study by Wharton Human-AI Research (WHAIR) and research firm GBK Collective surveying 1,007 U.S. enterprise leaders, the single largest barrier to stronger AI returns is the cost of tools, compute, and licenses, cited by 34% of respondents. Nearly two-thirds of surveyed leaders identified rising token, license, and API charges as a major obstacle to expanding AI usage—a figure that rises to 74% among vice presidents and C-suite executives.

The pressure on enterprise customers mirrors massive capital commitments on the vendor side. According to Reuters, Gartner analyst Will Sommer projected that AI providers will spend nearly $1 trillion on AI infrastructure in 2026 and an additional $6 trillion between 2027 and 2030, noting that current enterprise and consumer revenue streams may not be sufficient to yield profitable returns on all planned capital expenditure. As vendors seek to recoup those outlays, enterprise buyers face sustained pricing pressure. To manage these variable expenses, Javelin Strategy & Research notes that financial institutions are increasingly applying FinOps—financial operations frameworks originally developed to track and optimize variable cloud computing consumption—directly to their AI workloads.

AI Spending Moves Outside Traditional IT Budgets

The scale of capital flowing into AI has fundamentally altered corporate budget ownership. According to TechTarget, a BCG survey of more than 1,300 C-suite executives and senior leaders found that corporate AI spending doubled over the past year to reach 3.3% of total revenue (while a summary of BCG’s 2026 Applied AI Index by ZDNet notes that spending more than tripled from a 2025 baseline of 1% of revenue). Crucially, the BCG data reveals that 80% of corporate AI spending now sits outside the conventional enterprise IT budget.

Even as costs strain departmental budgets, allocations continue to climb. The Wharton and GBK Collective survey found that 87% of U.S. enterprise leaders expect their AI budgets to increase over the next 12 months, with 47% projecting budget growth above 10%. Similarly, a KPMG Global AI Pulse survey of 2,131 senior leaders across 20 countries, reported by PureAI, found that average planned corporate AI investment for the coming 12 months reached $210 million, up from $186 million in the first quarter of 2026.

With business units funding their own AI software and API consumption, Chief Information Officers (CIOs) and Chief Financial Officers (CFOs) are being forced to redefine decision rights. According to TechTarget, an IBM Institute for Business Value and Oxford Economics study of 1,500 senior finance leaders in September 2026 found that 62% of CFOs have taken on enterprise technology or AI strategy leadership responsibilities, 56% hold greater capital reallocation authority, and 54% are more involved in business model design.

The ROI Gap: Tracking Efficiency Versus Revenue Growth

While companies are spending heavily, industry surveys show conflicting realities regarding how far enterprises have progressed and whether those investments are paying off. According to ZDNet’s coverage of the BCG 2026 Applied AI Index, 50% of surveyed companies now generate value from AI and are transitioning from experimentation to production deployment, with only 4.5% of firms stagnating in 2026 compared to 14% in 2025. By contrast, a separate enterprise maturity survey by Further reported by Yahoo Finance found that 69% of organizations investing in AI remain in the exploring, experimenting, or implementing stages.

A similar divergence appears when measuring financial returns against total expenditure. In the Wharton and GBK Collective study, 80% of U.S. enterprise leaders reported positive returns on their AI investments, yet only 28% described those returns as strongly positive. Meanwhile, the KPMG survey found that while 89% of senior leaders reported measurable AI benefits, only 12% consistently assessed AI value against costs across their entire organization. Instead, KPMG found that companies typically review costs at project approval (61%) and after systems begin operating (59%).

The Wharton study points to metric selection as a primary differentiator between modest and strong returns. Across enterprise departments, “hours saved” remains the most common return-on-investment (ROI) metric at 40%, whereas net-new revenue is the least common major metric at 27%. However, among departments that track at least one growth metric—such as net-new revenue, sales conversion, output volume, or customer satisfaction—35% report strongly positive returns, compared to just 13% among departments measuring efficiency alone.

“Companies believe AI is going to transform their industries, but many are still asking it to prove itself by saving time,” Prasanna “Sonny” Tambe, professor of operations, information and decisions at Wharton and faculty co-director of WHAIR, told Yahoo Finance. “The companies seeing stronger returns are going further. They are redesigning workflows, putting clear ownership and governance in place, and measuring AI against outcomes that matter to growth.”

BCG’s findings align with this concentration of effort: 61% of AI-mature companies have moved away from isolated pilots in favor of multi-year enterprise programs, generating 2.4 times more top-line growth than laggards. BCG found that 54% to 78% of AI value is concentrated in core business functions—led by IT, research and development, marketing, sales, and customer service—and noted that agentic AI accounted for 22% of total AI value in 2026, up from 17% in 2025.

Frequently Asked Questions

How are companies dividing AI budget oversight when spending sits outside IT?

Enterprises are forming joint governance councils and splitting consumption costs across departments. According to TechTarget, companies are dividing responsibilities into three distinct lanes: business unit leaders own the workflow changes and business outcomes; the CIO manages the architecture, data security, integration, and governance; and the CFO controls capital allocation and evaluates whether projects should be scaled, redirected, or shut down. At online accounting software firm Xero, CFO Claire Bramley noted that the company centralized AI oversight in early 2026 under a leadership team that includes the CFO, chief legal officer, and global chief strategy officer. Other firms divide annual AI consumption expenses directly between product-and-technology teams and the central IT budget.

Does appointing a Chief AI Officer improve financial returns?

Survey data indicates a strong correlation between dedicated executive ownership and higher reported returns, though researchers note this may partially reflect higher baseline organizational commitment. In the Wharton and GBK Collective study, 58% of leaders at organizations with a newly designated Chief AI Officer reported strongly positive ROI. When AI oversight was simply added to an existing executive’s duties, strongly positive returns dropped to 24%, and fell to 8% at organizations with no dedicated AI leadership role.

Sources

Read More