From AI projects to profits: How agentic AI can sustain financial returns

This latest perspective on generative AI seems to reflect a maturation in how organizations approach implementing various forms of the technology and their business impact. To learn more about these evolving dynamics in AI, including the latest focus on agentic AI, the IBM Institute for Business Value (IBM IBV) fielded two executive surveys in partnership with Oxford Economics. Our research found three distinct trends related to how enterprises are approaching this most recent stage in their AI journeys.

First, our research reveals a new way of thinking about AI as organizations reconcile ambitious expectations with practical realities—including a shift from short-term, project-based ROI measurements to what actually impacts the company’s bottom line.

Second, we explore the emergence of agentic AI—autonomous systems capable of orchestrating complex workflows with a pertinent, personalized human partnership—the latest variation of AI technology and its impact on business and society.

Finally, we document diverging paths between organizations that implement AI with robust capabilities—demonstrating superior outcomes across AI-related revenue growth, operating profits, and customer satisfaction—compared to those pursuing fragmented approaches.

The AI ROI reset: A shifting focus from projects to profit

In the heady early days of generative AI, the technology seemed poised to deliver an economic windfall with minimal effort. In 2023, corporate executives rushed to implement pilots that promised—and initially delivered—eye-watering returns of 31%. But even as the champagne corks were popping, a more complex reality was evolving.

Enterprises are no longer reporting stratospheric ROI from generative AI pilots; as pilots scale, those returns have fallen back to earth. Projects that once boasted spectacular short-term gains have settled into a more pedestrian 7% ROI—notably shy of the approximately 10% cost of capital that serves as a typical capex hurdle rate. Over the past three years, CEOs say only 25% of AI initiatives have delivered expected ROI, increasing the pressure on business and technology leaders to demonstrate financial impact.

However, the top decile of organizations has achieved ROI of approximately 18%—well above the cost of capital—so higher returns are indeed achievable. More tellingly, organizations report sustained growth in operating profit improvements attributed to AI since 2022 (see figure). That’s not just a theoretical business case brought to a CFO to greenlight a project, but bottom-line impact delivered to shareholders.
 

AI’s impact on operating profit demonstrates an upward trend.

Organizations also are redirecting their AI investments toward core functions, which now command 64% of AI budgets compared to 36% for noncore activities. This reallocation suggests a growing sophistication: a recognition that AI delivers its most compelling value when applied to central business operations rather than peripheral processes. 
 

Working at the core is much more complicated than grabbing low-hanging fruit around the periphery, which may help explain why the pivot to core functions is concurrent with a decrease in ROI. Long term, the core should deliver far more scale and sustainable returns, but it requires more coordination to get moving.

Across organizations, experience has brought wisdom. Only 6% now pursue AI in an ad hoc fashion, down from 19% a year earlier. This shift reflects a broader trend toward strategic implementation across horizontal functions, industry vertical workflows, products and services, and even business models. Companies have begun to recognize that scattered pilots, while sometimes instructive, cannot deliver the systematic benefits that come from coordinated, enterprise-wide approaches and trusted data.

Yet for all this progress, more opportunity remains. Fewer than a quarter of organizations are reimagining their workflows with AI at the center, as a core engine for growth in their products, or fundamentally reimagining their business models.
 

This cautious incrementalism may represent a significant missed opportunity. AI's greatest potential lies not in making existing processes marginally more efficient, but in enabling entirely new approaches to value creation and delivery.

For now, the AI revolution continues its march not with the drama of dizzying early predictions but with the steady determination of a technology finding its proper place in the corporate arsenal—less miraculous perhaps, but ultimately more meaningful. The advent of agentic AI, as discussed in the next section, is a significant step in that direction.

Read the report to do a deeper dive into these three areas, as well as access critical actions that you can integrate into your organization’s AI strategy.