New Delhi

As generative artificial intelligence rapidly transforms how organisations operate, a widening gap has emerged between its potential and real-world outcomes, according to Tata Consultancy Services CEO and Managing Director K Krithivasan.

In an article published on the World Economic Forum website, Krithivasan cited research from MIT showing that nearly 95 per cent of enterprise AI pilot projects have failed to deliver tangible business value, despite widespread investment and optimism around the technology.

He noted that GenAI marks a structural shift in how organisations think, decide and act, requiring leaders to redesign not just individual decisions but the very architecture of decision-making. Success, he argued, depends on reimagining collaboration between humans and AI systems, rather than treating AI as a standalone tool.

To translate AI’s promise into performance, Krithivasan stressed the need for new organisational frameworks that enable continuous, informed engagement with AI while ensuring meaningful human oversight. Looking ahead to 2026, he said a new form of organisational intelligence is taking shape, where humans and machines jointly influence how options are generated, evaluated and debated.

According to him, realising AI’s value requires ecosystem-wide collaboration, strong domain expertise, robust governance and responsible scaling through co-innovation. AI, he said, should ultimately strengthen decision-making by offering data-backed choices rather than replacing human judgment.

Krithivasan outlined five guiding principles to build this foundation. Trust, he said, must be cultivated gradually through iterative implementation that allows stakeholders to gain confidence in how decisions are structured. Visibility into high-impact decisions is equally critical, with accurate and comprehensive data forming the backbone of effective AI-driven systems.

He also emphasised the importance of open-minded organisational cultures that allow intuition to be challenged and mistakes to be acknowledged. Traditional decision-making hierarchies, he added, must evolve so that insights from AI systems are not ignored due to legacy power structures.

ALSO READ:  India to construct a mega project for endangered one-horn Rhinoceros in Kanziranga

Finally, Krithivasan underlined the need to redesign workflows to support human-machine collaboration. This shift, he said, is less about AI literacy and more about building internal systems that make AI-generated insights clear, accessible and actionable across the organisation.