Enterprise AI is advancing into a more critical stage. The initial phase witnessed rapid deployment of pilots and copilots, as organizations hastened to demonstrate progress. Now, however, the focus has shifted to a fundamental question: how much of this experimentation is translating into tangible business value?
This inquiry highlights a growing disparity. Creating a pilot is relatively straightforward; however, integrating AI into the core operations of a business presents a far more significant challenge. The issue has transitioned from access to models or tools to organizations’ readiness to reassess processes, bolster infrastructure, and measure value with greater rigor.
Kenny Kesar, Global CIO of Wipro Limited, emphasizes that the core bottleneck lies not in the pilot phase but in the overall process. He observes that many organizations confuse early movement with true transformation, stating, “It’s very easy to create pilots.” The real hurdle emerges when attempting to transition from pilot projects to full-scale production. Kesar points out that many initiatives falter not due to technical shortcomings of AI but because businesses treat it as an auxiliary feature rather than fundamentally redesigning existing processes. “Most of the things are failing not because AI failed, it’s because the scenario in which we used AI made it very expensive,” he explains.
This insight addresses a critical dilemma. Many organizations anticipate substantial gains yet apply AI to workflows without rethinking them for increased autonomy, speed, or informed decision-making. Consequently, the outcome is frequently low returns on high investments. Thus, the gap extends beyond pilot versus production; it encompasses the chasm between innovation and operational model transformation.
Complementing this is the notion of readiness. Gajanan Palsule, Chief Architect for GenAI Enablement at TCS, characterizes much of the initial rush as “makeup AI” or “lipstick AI,” referring to efforts more aimed at showcasing relevance than resolving significant challenges. His assertion underscores the fact that many enterprises remain neither cloud-native nor data-ready, attempting to scale AI atop outdated, monolithic systems. While this may suffice for small pilots, such infrastructure fails to support expansive organizational use cases.
This is why discussions around data readiness, system architecture, and adaptable design continue to occupy central importance. AI may appear to occupy an apex position, but its success relies heavily on the foundational elements that support it. A solid infrastructure may harbor success for limited experiments but often erodes under enterprise scale.
Deepti Shibad, Director of Tech Strategy and Digital Technology Transformation at CRISIL Limited, articulates that the true measure of maturity begins when AI performance is gauged not by activity levels but by outcomes produced. She identifies challenges across three dimensions: people, processes, and technology. Skillsets require evolution, delivery models need transformation, and the foundational infrastructure—cloud, security, integration, and data—must be robust enough to accommodate scaling.
While pilots may function effectively in isolated environments, the same cannot be said for production systems. Additionally, AI initiatives must remain aligned with business strategies. Raman Srinivasan, CDO of InMobi, stresses this alignment, noting that AI must connect directly to goals of growth, efficiency, compliance, or innovation from its inception. An absence of this coherence may lead to projects that, despite continuing operation, lose clarity of purpose.
The overarching lesson is clear: enterprise AI is moving beyond its initial novelty and into a phase of accountability. Organizations that distinguish themselves will not be those that launched numerous pilots or proclaimed transformational efforts. Rather, they will be the ones that reengineered processes, strengthened technical foundations, and maintained a focus on measurable business outcomes.
This convergence of promise and execution is where enterprise AI will ultimately reveal its true potential or stagnate.
Disclaimer: The views expressed are solely those of the speakers and have been derived from the ETCIO Cloud Summit 2026. ETCIO does not necessarily endorse them.
With contributions from Swati Sengupta.
Published on April 17, 2026, at 08:50 AM IST
