Kumar asserts that the disparity between investment and value generation is the primary obstacle as the industry transitions into 2026.
“If enterprises want to realize value from this level of expenditure, they should be producing trillions in business value. Infrastructure providers themselves require $600-700 billion in returns for even a 10% success rate. That expected trillions in enterprise value has yet to materialize,” Kumar stated in an interview with CNBC-TV18.
He elaborated that the industry is currently engaged in the initial stages of the AI stack—focusing on compute development and model training access—and is just starting to progress toward creating applications that will transform this infrastructure into financial results.
Adoption Lagging Behind Infrastructure Buildout
Kumar pointed out a noticeable discrepancy in progress:
AI infrastructure is advancing at an exceptionally fast pace, while adoption and integration within enterprises are lagging.
This delay, he emphasized, has been observed in earlier technological cycles as well. However, the current urgency is heightened due to AI’s unique nature: its scaling laws evolve within six to twelve months, much more rapidly than the decades-long cycles dictated by Moore’s Law.
“The longevity of AI infrastructure is limited. If enterprises fail to derive value swiftly, the technology risks becoming obsolete,” he cautioned.
The Productivity Curve Will Take 10-20 Years
Kumar envisions that significant productivity enhancements will follow a lengthy trajectory—similar to the internet’s evolution from a peripheral aspect of business in the 1990s to its current prominence in everyday life.
He noted that enterprise productivity will be the initial focus, followed by broader societal productivity. Nonetheless, this shift will not occur overnight.
“The journey toward productivity spans 10 to 20 years. Each year will bring improvements as tasks become automated, workflows are integrated, processes are revamped, and ultimately, business models are restructured,” he added.
Also Read | AI stocks in bubble zone, valuations mirror 1999-2000 dot-com mania: Manish Chokhani warns
Cognizant’s Role: The ‘AI Builder on the Bridge’
Kumar has characterized Cognizant as an essential connector between creators of AI infrastructure and enterprise users.
“We refer to ourselves as AI builder companies on the bridge,” he explained, highlighting the function of technology services firms in converting infrastructure capabilities into genuine business applications.
The Road to 2026
As the industry looks toward 2026, two key themes are expected to emerge, according to Kumar:
- Machine-assisted software cycles becoming commonplace, and
- Agentic AI transforming operations, creating new opportunities for IT firms.
However, for enterprises, the message is unequivocal: the billions invested in 2025 will only be justified if they expedite tangible AI adoption in the forthcoming years.
Also Read | Why India’s limited AI exposure could be a hidden advantage for global investors: BlackRock’s Ben Powell
Watch the accompanying video for the complete conversation.
