When bigger AI budgets don’t mean better results

Enterprise AI spending surges while ROI remains elusive

Generative AI spending has exploded across industries. Boards are demanding visible progress, and most enterprises have moved from pilot projects into full-scale AI integrations. Yet measurable returns remain inconsistent. Forrester Research reports that budgets for generative AI rose sharply over the past year, but most companies still struggle to show lasting ROI. Early pilots display quick gains in automation and efficiency, but those results often fade when scaled into real-world operations. Costs rise unexpectedly, benefits remain difficult to quantify, and governance requirements make financial justification more complex.

This is not a failure of technology. It’s a failure of traditional measurement models. Generative AI doesn’t fit cleanly into standard budgeting frameworks because it behaves differently, its costs depend on use, and its benefits are often indirect or long-term. Many organizations track AI spending as if it were traditional IT infrastructure, but AI’s economics are fluid and consumption-driven. That mismatch destroys clarity around value creation.

C-suite leaders should stop expecting AI to prove its worth through old financial models. Instead, they should focus on defining value in context. That means quantifying productivity impact, decision quality, and speed-to-market improvements, not just cost savings. AI can’t be measured by transactional outcomes alone. It changes how companies operate and compete. Without adapting financial expectations to that reality, ROI will always appear elusive, even when strategic value is being created.

Transition of IT finance from cost control to strategic value creation

Enterprise finance needs to evolve. The organizations seeing success with AI are those that treat IT budgets not as costs to be minimized but as strategic engines for growth. Greg Zorella, Lead Principal Analyst at Forrester Research, explains that high-performing companies link every technology dollar to strategic results, customer expansion, market share growth, and efficiency improvements. This shift is key to understanding AI’s economic logic. Traditional IT investments are predictable and easy to depreciate over time. Generative AI isn’t. It’s dynamic. Consumption patterns shift constantly, making fixed models unreliable.

For executive teams, this means that finance and IT must work together in new ways. Cost transparency and shared attribution models become essential. Everyone, from IT to marketing, must agree on how value is defined and tracked. That collaboration creates accountability and data-driven decision-making. Zorella also notes that overspending is not automatically negative if the money fuels high-value outcomes. The real failure is spending without prioritization, funding projects without understanding tradeoffs or potential returns.

Executives should embrace a value-oriented finance mindset. Instead of limiting AI through rigid cost controls, they should give teams space to invest where returns are provable and aligned with business goals. The challenge is operational, not philosophical. Leaders must build financial systems that evolve as consumption changes, ensuring AI investments are continually tied to measurable outcomes. This mindset will help enterprises move from AI experimentation to AI-driven strategy.

Corporate budget realities demand tangible business outcomes from AI initiatives

AI spending has entered a new phase, measured, competitive, and accountable. The days of unrestricted AI budgets are over. Executives and boards now expect clear, quantifiable business results tied directly to company growth and efficiency. Sumit Johar, Chief Information Officer at BlackLine, explains that finance leaders no longer value broad adoption metrics. Stating that “95% of employees are using AI” is meaningless if it doesn’t lead to increased profitability, lower risk, or stronger revenue performance.

Johar distinguishes between two types of AI investments. The first involves “everyday AI”—tools that improve basic productivity such as writing, research, or data summarization. These can reshape how people work, but their impact on financial outcomes is difficult to prove. The second category consists of outcome-driven AI projects tied explicitly to business priorities, accelerating customer onboarding, improving sales forecasts, cutting operational costs, or speeding up deployments. These projects have measurable and defensible value.

Under today’s financial conditions, AI funding is no longer additive. CIOs must reallocate from existing budgets rather than request new ones. Johar notes that BlackLine already enforces this discipline. Projects are reviewed by IT, finance, and business leaders together, ensuring accountability and alignment. For executives, the key message is simple: AI cannot remain an isolated investment. It must produce verifiable returns that hold up under financial scrutiny. That mindset turns AI from experimental spending into a core mechanism for competitive advantage.

Structural challenges in scaling AI lead to deteriorating ROI

Enterprises often see early AI success in controlled, small-scale environments. But once projects scale into production, costs explode and ROI declines. Jim Olsen, Chief Technology Officer at ModelOp, describes this as a structural challenge, not a technical one. Initial AI deployments operate with clear parameters, limited data, and manageable costs. When exposed to real-world complexity, usage patterns change unpredictably, and consumption costs rise faster than expected.

Generative AI intensifies this problem. Because it depends on free-form user interaction, consumption rates, measured in tokens or computation, can shift dramatically. As more teams reuse models across departments, the ability to assign costs or attribute value to specific outcomes breaks down. Without lifecycle tracking and an accurate inventory of what systems are running in production, organizations lose control of both cost and measurement. According to Olsen, many enterprises cannot even identify all the AI systems currently in use, a fundamental barrier to governance and accountability.

For C-suite leaders, the message is direct: operational discipline must replace ad hoc experimentation. AI requires structured lifecycle management, covering development, deployment, monitoring, and retirement. This approach provides clear ownership and measurement frameworks, allowing decision-makers to manage cost exposure as systems scale. Without it, pilot projects that once looked promising deteriorate into expensive, unmeasurable expenses. Treating AI as continuously managed infrastructure ensures that ROI remains visible and defensible even under expansion pressure.

Effective governance is key to sustaining and defending AI ROI

Governance now defines whether AI investments succeed or fail. As AI systems become more integrated across business operations, executives are realizing that performance and compliance alone are not enough. The ability to prove value and manage accountability determines long-term ROI. Anthony Habayeb, Chief Executive Officer and Co-founder of Monitaur, points out that many companies struggle with ROI discussions because they never set clear definitions of success from the start. When outcomes and metrics remain undefined, organizations are left defending expenses after the fact, often without the documentation needed to justify the original investment.

Strong governance changes that dynamic. It introduces structure early in the lifecycle, before deployment, through defined objectives, transparent data tracking, and continuous validation. Habayeb emphasizes that governance should not be viewed as a bureaucratic constraint or mere compliance mechanism. Instead, it’s a framework that enhances operational clarity, surfaces optimization opportunities, and limits waste. Well-governed AI systems enable leadership teams to identify inefficiencies and improve both accuracy and performance.

For C-suite executives, this level of governance is becoming mandatory. Regulatory shifts are adding urgency. Frameworks such as the EU AI Act are already influencing global standards for transparency, risk management, and auditability. Organizations that establish mature governance early will not only meet compliance demands but also gain a competitive advantage. They can demonstrate ROI with confidence and defend AI decisions before shareholders, regulators, and customers. Habayeb summarizes it clearly: “If you don’t know what success looks like at inception, you can’t defend ROI later.”

AI adoption is maturing into an accountable, Outcome-Centric enterprise asset

Enterprise AI is moving past the experimental stage and into a state of measurable accountability. The companies leading in AI adoption are no longer focused solely on innovation, they are focused on sustained, provable business value. They understand that AI’s potential depends on integration with business strategy, aligned budgets, disciplined operations, and transparent governance. When these elements work together, AI becomes an asset with visible contribution to business outcomes, not just a line item in an innovation budget.

This new phase requires executives to take ownership of AI value creation. Financial models must be adjusted to fit usage-based economics, and value definitions must extend beyond direct profits to include efficiency and risk reduction. Operational discipline must stay consistent across the model lifecycle, enforcing visibility from build to retirement. Governance should remain embedded, not added later. These practices ensure that AI decisions are strategic and data-backed.

For decision-makers, this shift represents an important lesson. AI will not justify itself; its value must be designed, measured, and validated continuously. The organizations succeeding now are those that treat AI as a long-term business system, not an experiment or short-term showcase. By connecting financial clarity, lifecycle accountability, and proactive governance, enterprises are transitioning from trial-led enthusiasm to performance-driven endurance. The result is an AI strategy that can withstand scrutiny, scale sustainably, and deliver tangible competitive advantage.

Key takeaways for decision-makers

• AI spending rises faster than ROI clarity: AI budgets continue to soar, yet most organizations fail to prove lasting financial impact. Leaders should redefine ROI metrics to capture productivity, efficiency, and strategic gains beyond traditional cost savings.
• Finance must evolve from control to value creation: Legacy budgeting models can’t measure AI’s unpredictable, consumption-driven costs. Executives should align finance and IT around shared attribution models that link AI investments directly to growth and competitiveness.
• Budgets now demand clear business outcomes: AI projects must compete for funding based on tangible results, not experimentation. Leaders should prioritize initiatives that directly affect revenue, profitability, or risk reduction and enforce joint accountability between IT and business teams.
• Scaling AI without structure erodes ROI: Early AI pilots perform well, but costs escalate as projects move into production. Decision-makers should invest in lifecycle management and usage tracking to control cost volatility and preserve operational efficiency.
• Governance defines ROI defensibility: Weak governance undermines ROI justification even for well-performing AI systems. Executives should embed governance early, integrating clear objectives, reporting, and monitoring to strengthen accountability and regulatory readiness.
• AI maturity depends on measurable accountability: The experimental phase of AI is ending, replaced by outcome-driven oversight. Leaders should treat AI as a strategic, accountable asset, aligning it with core business goals, sustainable cost models, and transparent governance frameworks.