India AI Summit: Biggest AI opportunity lies in bridging gap between LLMs and enterprises, says Vishal Sikka

Vishal Sikka

• AI's biggest value lies in bridging LLMs and business users
• Trust, reliability, and usability limit AI's enterprise impact
• Experts emphasize safer, efficient, and reliable AI development

Did our AI summary help?

The biggest value from artificial intelligence (AI) will come not from building ever-larger language models but by bridging the gap between LLMs and real business users, former Infosys CEO and founder of Vianai Systems Vishal Sikka said on February 19.

Speaking at the India AI Impact Summit 2026, Sikka said AI could deliver dramatic productivity gains for users who understand how to work with it but its impact remains uneven and limited by trust, reliability and usability issues within organisations.

“People who understand how to use AI are astonishingly effective with it,” Sikka said, citing examples where decisions that earlier took months or even a year were completed in a matter of days using AI-driven simulations.

He, however, cautioned the effectiveness is not uniform, describing AI’s progress as a “jagged frontier”.

These comments assume significance as global markets have been unsettled by the latest releases from Anthropic, with new Claude tools triggering sharp valuation resets across several software and services companies, as investors reassess how quickly AI can displace existing business models.

According to Sikka, the real opportunity lies in building systems that sit above language models and translate their capabilities into trusted, verifiable, and reliable outcomes for business users. Enterprises face a wide gap between what LLMs can do and how value is actually delivered on the ground, he said.

“Overcoming that gap is where a lot of value-creating opportunities lie,” he said, adding without correctness and trust, AI adoption in enterprises will remain constrained.

AI systems still suffer from problems such as hallucinations, safety risks, and very high-energy use. India should go beyond simply adopting current AI and focus on building safer and more efficient systems, Sikka said.

His comments echo concerns flagged by Infosys chairman Nandan Nilekani. Speaking at the recent Infosys AI Day, Nilekani said while AI models have advanced rapidly, enterprise adoption lagged behind. Execution, and not opportunity, is the main risk for businesses, he added.

Cognizant CEO Ravi Kumar S recently said orchestration of AI would need IT firms and will not take place "magically".

These remarks point to a converging view among industry leaders that while large language models have unlocked unprecedented capabilities, the harder challenge now lies in making AI work reliably at scale inside enterprises, where legacy systems, regulatory requirements and real-world risks continue to shape adoption.