TCS CEO on GenAI: Bridging the Gap Between AI Promise and Performance

Generative AI shows promise, but 95% of enterprise AI pilots fail, says a study cited by TCS's CEO. To unlock AI's value, leaders must rethink decision-making, design new frameworks, and foster effective human-AI collaboration.

As generative artificial intelligence (GenAI) reshapes how organisations think and act, a stark gap has emerged between promise and performance. The rise of GenAI represents a fundamental shift. It transforms how enterprises think, make decisions and act. Citing a study by MIT, K Krithivasan, CEO & Managing Director, Tata Consultancy Services, said in an article on the World Economic Forum (WEF) website that despite the enormous potential, research shows that 95% of enterprise AI pilots have failed to deliver measurable value.

A New Framework for AI Success

To move beyond the promise of AI and turn potential into performance, leaders must rethink not only what decisions are made, but how decision-making itself is designed, and how humans and AI can collaborate most effectively. For success to be achieved and AI's potential to be realised, new frameworks are required. Leaders must establish structures that enable continuous, informed dialogue with AI, while keeping humans in the loop, he said.

"As we look ahead to 2026, a clearer picture of AI's impact is emerging. We are witnessing the advent of a new form of organisational intelligence, where combinations of humans and machines shape how choices are developed, presented and discussed," the article read.

He said that to unlock AI's true value, businesses need to collaborate across ecosystems, apply deep domain insights, ensure rigorous governance and scale responsibly through co-innovation models. AI should improve organisational decision-making by presenting better choices backed by data.

Five Core Principles to Build a New Foundation

The CEO & MD highlighted five core principles to build this new foundation.

Build Trust Over Time

"Trust must be built over time." He said that to manage risk and improve trust, implementation must be a gradual process. Stakeholders need to feel growing confidence in how decision environments are framed, not just in the correctness of decisions. "Taking an iterative process that allows for incremental learning can help colleagues understand how to operate within new architectures successfully," he added.

Ensure Visibility and Accurate Data

"Visibility is key." Organisations must track how high-stakes decisions are made and take a holistic view of their business and the wider landscape. Intelligent Choice Architectures are dependent on accurate data and will fail - or even model the wrong choices - without it.

Foster a Culture of Open-Mindedness

"Open-mindedness is essential." Intelligent Choice Architectures don't flatter intuition; in fact, they often challenge it. Organisations must build a culture where people are comfortable about getting things wrong in front of others. In the shifts to harnessing these systems, this is part of the journey to success.

Evolve Decision-Making Hierarchies

"Decision-making hierarchies must evolve." To unlock true value from AI agents, decision-making-related barriers need to be removed. If only credentialled experts of legacy hierarchies can make decisions, Intelligent Choice Architecture insights may be dismissed, regardless of their quality.

Adapt Workflows for New Insights

"Workflows must change." If cultural shifts can be achieved, new workflows must also be developed to support them. Processes must be more flexible to allow better options presented by Intelligent Choice Architectures to be implemented. This is not a question of AI literacy, but of building internal systems that make these human-machine-cooperation-derived insights visible, understandable, and actionable.

(Except for the headline, this story has not been edited by Asianet Newsable English staff and is published from a syndicated feed.)