Artificial intelligence has moved from research laboratories to everyday tools at a remarkable speed. Yet, for many organisations, the promise of AI remains frustratingly out of reach as a large share of AI initiatives never move beyond pilot stages.
According to an MIT report published last year, 95% of AI pilot programmes deliver little to no measurable impact on companies’ P&L. It blamed flawed enterprise integration.
However, Rohit Tandon, Senior Managing Director and Global Head of AI and Data at Deloitte Consulting, believes the biggest obstacle is not technological capability but organisational mindset.
“The biggest bottleneck is culture,” he tells AIM, explaining that companies often struggle to overcome internal resistance to change even when the technology itself is ready.
The paradox is striking. Artificial intelligence has already become mainstream across global business. According to McKinsey & Company’s latest State of AI survey, 88% of organisations now use AI in at least one business function—a dramatic increase from 55% just two years earlier.
But adoption does not necessarily translate into transformation. The same survey found that only about 38% of companies have successfully scaled AI beyond pilot projects, leaving the majority stuck in experimentation mode.
That gap between enthusiasm and impact reflects a deeper organisational challenge.
The AI Pilot Trap
For Tandon, the reason is straightforward: organisations often treat AI as a technology experiment rather than a business transformation opportunity.
Teams may spend months testing capabilities that have already been widely proven elsewhere—such as speech recognition or automated content generation—instead of focusing on how AI can solve real operational problems.
“Why do you need to experiment whether AI can take voice and convert it to text?” he ponders.
The challenge is widely recognised across industry research. A 2025 World Bank digital technology report noted that AI adoption remains heavily concentrated in large firms and knowledge-intensive sectors, with many smaller organisations struggling to move from experimentation to deployment.
For companies serious about transformation, Tandon argues that the starting point should not be the technology itself but the business problem.
“You might spend ₹100 and not get ₹1,000 back,” he says. “But maybe you get more customers, or your customer is happier.”
In many cases, the real value of AI is realised through improved customer experience, faster decision-making, or entirely new services.
The ‘Tick-Box’ AI Race
Another dynamic shaping corporate AI strategies is the pressure to demonstrate technological progress.
In a market where AI has become the defining technology narrative, companies often feel compelled to show investors that they are actively investing in it. That has led to a proliferation of announcements highlighting AI agents, copilots, and experimental deployments.
Tandon describes this as a “tick-box” approach to AI.
“I don’t want to call it a rat race,” he says. “But everybody wants to say ‘we are using AI’.”
The problem is that announcements do not always translate into results. Companies may showcase hundreds of AI pilots without delivering measurable business outcomes.
Industry data suggests the scepticism is justified. McKinsey State of AI survey shows that only a small minority of organisations report significant financial impact from their AI deployments—even as adoption becomes nearly universal.
The disconnect highlights a broader challenge for companies navigating the AI boom: balancing the pressure to signal innovation with the slower, more complex process of transformation.
Growth—AI’s Biggest Opportunity
A final misconception shaping corporate strategies lies in how executives frame the role of artificial intelligence. For many organisations, AI is still viewed primarily as a cost-cutting tool. Automation and efficiency gains dominate early deployments.
“About 90% of CEOs or CXOs look at it as a cost lever,” Tandon reveals.
Yet the technology’s real potential may lie in enabling entirely new products, services, and customer experiences.
Few sectors illustrate this shift more clearly than IT services, where many traditional outsourcing roles involve rule-based processes such as system monitoring and troubleshooting.
“A lot of the work in that space is rule-based and mechanical,” Tandon says. “High time that all of that gets automated.”
In India, where IT services form a major pillar of the technology economy, the implications could be significant. But technological transitions rarely eliminate work. When the IT industry first expanded decades ago, similar fears of job losses emerged. In practice, automation removed certain roles but also created entirely new ones.
The Compute Revolution
The recent explosion of interest in generative AI can be directly tied to the growing computing power.
Advances in large-scale infrastructure have allowed companies to run far more complex AI systems than earlier generations of machine learning could support. Instead of relying on a single model, modern AI systems often combine multiple models working together, interpreting queries, retrieving information, and refining outputs.
“The fundamentals of AI are the same,” Tandon says, pointing instead to the ability to run algorithms at unprecedented scale.
That scale has also expanded AI’s reach far beyond enterprise software. According to research from the Microsoft AI Economy Institute, roughly 16% of the global population—about one in six people—now uses AI tools to learn, work, or solve problems, highlighting how quickly the technology has entered everyday life.
AI is becoming cheaper, more powerful and more widely available with each passing year. Yet many organisations still struggle to integrate it meaningfully into their operations. The organisations that ultimately succeed in the AI era may not be those with the most advanced algorithms. They will be the ones willing to change.
Your reaction
Discussion
What's your take on this story?
No comments yet — be the first to share your view.