In 2025, a significant number of Indian and global enterprises embraced the surge toward generative AI, aiming to enhance automation and speed. However, as 2026 unfolds, many organizations are confronting a stark reality: AI’s capabilities, while impressive, often lack the depth required to navigate the complex dynamics of the Indian market. A recent study by MIT highlights that nearly 95% of generative AI projects are failing to yield a clear return on investment (ROI). This issue is particularly pronounced in India, where supply chains must contend with multi-modal infrastructure challenges and fluctuating seasonal demands, underscoring what some analysts refer to as the “pilot trap.”
The central issue lies in the distinction between data and knowledge. Indian enterprises have made strides in digitizing their operations and consolidating data, yet mere integration does not ensure a shared understanding among varied functions. Technical connectivity is merely a starting point; without semantic alignment, AI systems often interpret fractured signals, which can lead to misguided decisions.
Moreover, around 70% of enterprise knowledge remains undigitized, often referred to as “tribal knowledge.” While AI can process transactions, it struggles to comprehend human intention. Progress requires a robust knowledge framework that links structured data with the cognitive reasoning essential to business functions.
Agentic AI heralds a promise of autonomy, but achieving coherence amidst that autonomy is crucial. For instance, in a supply chain scenario, an AI agent could autonomously reroute a shipment for cost efficiency, but this decision might disregard impacts on customer deadlines or the organization’s capital objectives—rendering the effort ineffective. For AI to be functional across various departments, including commerce, finance, and procurement, a unified enterprise intent is necessary.
This calls for a contextual intelligence layer that integrates with existing enterprise resource planning (ERP) and customer relationship management (CRM) systems. Rather than supplanting these systems, this layer operates as a nervous system that harmonizes data and enables reasoning across the entire value chain. As coherence emerges across the organization, the competitive factor shifts from software capabilities to the architecture of integrated knowledge.
A significant transition in this paradigm is the role of AI agents: they are not meant to replace Indian professionals but to work cooperatively with them. In a country with a rich labor pool yet a shortage of specialized talent, these AI agents can manage the complexities of tracking numerous stock-keeping units (SKUs) and highlighting exceptions. This allows human planners to concentrate on high-value tasks, such as strategy, relationship management, and creative problem-solving.
We are moving toward an era of collaborative intelligence, wherein humans and systems synergistically reason together. As knowledge becomes more structured and accessible, the landscape of supply chain organizations is poised for transformation. Decision-making delays could diminish, hierarchies might flatten, and operators could achieve a level of visibility that previously necessitated frequent status updates. This scenario encapsulates workforce enhancement: a single planner can oversee a broader range of responsibilities confidently, fortified by a system that comprehends the drivers behind the data.
Looking ahead, the divide between market leaders and laggards in India is unlikely to hinge solely on who possesses the most cutting-edge algorithms. In a landscape where advanced AI models are quickly becoming commonplace, contextual understanding emerges as the real currency. Organizations that view AI as an ancillary feature for fragmented systems are likely to realize only minimal, localized improvements. Conversely, those that invest in a contextual knowledge backbone stand to unlock an adaptable, self-correcting enterprise model. In summary, Agentic AI lacking a knowledge framework amounts to rapid automation that is often disconnected and counterproductive. In contrast, Agentic AI anchored in context signifies coordinated intelligence, a distinction that will be critical in today’s volatile global market.
The author is Anand Srinivasan, Chief Strategy Officer of o9. The views expressed herein are solely those of the author and do not necessarily reflect the views of ETCIO. ETCIO disclaims any responsibility for damage caused to individuals or organizations directly or indirectly.
