Artificial Intelligence is entering the era of cost awareness as companies seek returns
New Delhi: The gold rush in artificial intelligence is giving way to a more disciplined phase. After two years of racing to deploy the biggest models and consume more compute, companies are asking a simpler question: Does every AI rupee generate measurable business returns? Uber recently admitted that internal AI coding budgets were exhausted much faster than expected as employee adoption increased, leading to tightened governance.
Meta has made reducing inference costs a strategic priority, while Amazon, Walmart, Cisco, and Uber have introduced usage limits or directed employees to cheaper AI models to contain costs. The pressure for accountability has become evident in India as well. An EY-CII survey found that 47% of Indian companies now have multiple generative AI applications in production, while more than 95% of them still keep AI and ML budgets below 20% of overall IT spending. A separate study by SAP found that Indian organizations expect investment in AI to rise by 45% over the next two years, even as the focus shifts towards improving returns rather than simply increasing deployments. “We are seeing diminishing returns more clearly,” said Sambhav Jain, managing director and partner at Boston Consulting Group in India. Larger context windows, multiple AI agents, and expensive heuristics often lead to only marginal improvements in the quality of output without corresponding gains in productivity, revenue, or customer experience, he said.
The problem is not that AI is spending on itself, but that it is spending without accountability, said Deepak Dhanak, co-founder and chief operating officer of Rocket, an AI-driven platform. “Overspending is not an AI problem, it is a measurement problem,” Danak said. “Token consumption became a trivial metric during the first wave of enterprise AI. Companies now need to match the right model to the right mission and measure outcomes, not activity.” This rethinking comes as artificial intelligence investments accelerate. Microsoft, Amazon, Alphabet and Meta were expected to spend nearly $320 billion on AI infrastructure in 2025. In India, a study by Z47-OpenAI-Zinnov found that nearly 90% of mature AI users have reduced some form of BPO spending, with more than a third reducing outsourcing work by more than 25%. The same study found that 86% of Indian startup founders plan to increase AI budgets this year, and more than half expect spending to more than double, yet only 9% have seen a measurable increase in sales or conversions attributable to AI.
“If you can’t draw a straight line from spending to value derived, you’re not measuring ROI — you’re measuring noise,” Danak said. Executives say companies are increasingly asking whether cheaper models can deliver 95% of the value for 20% of the cost. “We’re not experimenting with AI anymore — we’re running it. This shift is changing the entire cost conversation,” said Melesh J, Head of Strategy and Operations at SAP Labs India. “The real danger is not overspending on cryptocurrencies. It is mixing spending with strategy – and not having a measuring system to know which AI investments are actually paying off.”