The Soaring Cost of AI Success
Recent findings from Gartner have sent a clear message to the business world: achieving a positive return from artificial intelligence is becoming significantly more expensive. According to the research firm, organisations are now spending 18 times more to get
value from their AI initiatives compared to just two years ago. This 'AI inflation' isn't necessarily about the software itself getting more expensive. Instead, it reflects a deeper understanding of what it truly takes to make AI work in a corporate environment. The initial software or model subscription is just the tip of the iceberg. The real costs—and the bulk of this 18-fold increase—are found in the surrounding infrastructure, talent, data management, and governance required to move from a small-scale pilot to a value-generating, enterprise-grade system.
The Total Cost of Ownership Iceberg
This dramatic rise in spending highlights the total cost of ownership (TCO) for AI, which often surprises leaders. Early experiments might seem cheap, but production-level AI demands significant investment in several 'hidden' areas. These include readying and cleaning massive datasets, paying for the immense computing power needed for model inference (the process of an AI making a prediction), and integrating AI tools into existing workflows. Furthermore, there's the human element: hiring expensive AI specialists, retraining current staff, and managing the organisational change that comes with automation. Gartner notes that the costs of model inference alone will eclipse training costs, making up at least 70% of a model's lifetime costs by 2028. As one analyst put it, a cheaper AI model doesn't necessarily mean a cheaper business, because autonomous agents can turn operational delays into direct computing costs.
But 52% Are Seeing Positive Value
Here's the other side of the coin. In the same period of rising costs, a Q2 2026 Gartner survey found that 52% of over 1,000 functional leaders are already seeing 'some positive value' from their AI initiatives. This seems contradictory, but it's key to understanding the current AI landscape. The 'positive value' reported isn't always a hard financial return on investment. It often includes significant gains in productivity, improvements in customer satisfaction, faster idea generation, and enhanced operational efficiency. For many, these benefits are real and substantial, even if they don't immediately appear as a new revenue line on a spreadsheet. The data shows that while the price of entry is high, more than half of the companies making these investments are finding them worthwhile and are seeing tangible, positive outcomes that justify the spend.
Connecting High Costs to High Returns
The 18x cost increase and the 52% positive return are not opposing facts; they are two parts of the same story. Successfully deploying AI is an expensive, complex undertaking, and those who are seeing returns are the ones who are willing to make the necessary, holistic investment. The data suggests that success in AI is less about buying a tool and more about building a capability. Companies seeing the best returns are those redesigning their operating models and workflows around AI. A 2027 Gartner CEO survey found that 86% of chief executives identified this redesign of work as the top opportunity for generating value from AI. This is not about simply making old tasks faster; it's about fundamentally changing what work gets done and how a business competes. Those who treat AI as a strategic, transformative investment—covering everything from governance to talent—are the ones getting a return, even if that investment is far larger than it was two years ago.
The Road Ahead for Business Leaders
The key takeaway for leaders in India and beyond is to approach AI with a clear-eyed view of both the costs and the opportunities. The rapid pace of AI adoption is outpacing governance readiness, creating a risky environment where businesses are 'accelerating before they can steer'. In India and the APAC region, only 14% of CIOs have agentic AI governance capabilities in place, despite 74% planning to deploy such technology soon. Success in this new era requires a shift in mindset from rapid, isolated experiments to a deliberate strategy that aligns AI initiatives with business goals and risk management. This means being highly selective about which AI projects to fund and ensuring there's a clear path to value, however the organisation chooses to define it. The focus must move beyond simply adopting AI to ensuring it delivers sustainable, measurable outcomes.
















