Hook
Ravi Chauhan, CEO of Hexaware Technologies, told the press that AI could cut IT project costs by 25%. The market heard optimism. I heard something else entirely: a defensive admission that the traditional IT services business model is structurally insolvent.
"Solvency is not a metric; it is a moment of truth." And for an industry built on billing human hours, that moment has arrived.
The number itself isn't the story. Twenty-five percent aligns neatly with McKinsey's and Gartner's projections of 20-40% efficiency gains from AI across the software development lifecycle. The real signal is what the CEO didn't say: that the industry's core pricing mechanism—the billable human-hour—is being liquidated in real time.
Context
Let me position this properly. Hexaware is India's sixth-largest IT services firm, roughly 30,000 employees, annual revenue in the $1 billion range. It sits in an uncomfortable middle layer of the global IT services stack—below the tier-one giants (Accenture, TCS, Infosys, Cognizant) and above the AI-native startups that are attacking the same client budgets with fundamentally different cost structures.
The company was taken private by Carlyle Group in 2020 and returned to public markets in 2025. That timing matters. A newly listed mid-cap IT firm needs to project AI competence to hold its valuation. Chauhan's 25% figure is as much investor relations as it is operational forecasting.
But here's the uncomfortable structural reality: human labor constitutes 60-80% of IT project costs. When a CEO publicly acknowledges that AI can strip a quarter of that away, he's not announcing an efficiency gain. He's admitting that the industry's primary input—programmer hours—has been devalued at the margin.
Core
Let me run the forensic analysis on what this 25% actually means.
The pricing paradox. If AI cuts delivery costs by 25% but clients demand a corresponding price reduction—which they absolutely will, because the CEO just told them the number—then revenue falls proportionally. Cost down 25%, price down 25%, margins unchanged. The company runs faster to stay in the same place. This is the "efficiency trap" that IT services firms are walking into, and few have articulated a way out.
The client power shift. This is the ghost in the machine. Chauhan's statement confirms what clients already suspected: AI compresses delivery costs. The moment a service provider publicly validates that suspicion, the negotiation dynamic shifts. Clients now hold a 25% discount expectation as a baseline. The IT services firm isn't offering a concession; it's surrendering pricing power preemptively.
The competitive squeeze. Hexaware faces a pincer movement. From above, Accenture has committed $3 billion to AI, Infosys built the Topaz platform, Wipro allocated $1 billion. From below, AI-native tools like Devin and Cursor are attacking specific parts of the delivery stack with near-zero marginal cost. The mid-tier IT firm is caught between giants who can out-invest them and startups who can out-disrupt them.
The labor arbitrage collapse. Here's the uncomfortable historical context. India's IT services boom was built on one simple equation: a Chennai-based developer costs 20% of a Silicon Valley counterpart with 80% of the output quality. That arbitrage was the foundation of a $250 billion industry employing 5 million people. AI doesn't just erode that edge—it potentially eliminates it entirely. If an AI assistant can generate code at near-zero cost, the location of the human generating the prompt becomes nearly irrelevant.
The 25% figure, if realized, represents roughly $240-300 billion in annual value release across the $1.2 trillion global IT services market. The question no one is answering: where does that value go? To clients via price cuts? To shareholders via margin expansion? Or to AI infrastructure providers like OpenAI, Anthropic, and Microsoft who own the productivity layer?
Contrarian
Now let me push back against the prevailing narrative. Most analysts will frame this as "AI disruption hitting the IT services industry." I see the opposite. This is the IT services industry using AI as a survival mechanism to preserve relevance.
The contrarian view: IT services firms aren't being disrupted by AI—they're being rescued by it. Here's the logic. Their clients' software demands are growing faster than their ability to deliver. Backlogs stretch for years. AI provides the only lever to keep pace with demand without multiplying headcount. The 25% cost reduction is actually a capacity expansion story in disguise.
But here's the darker twist: AI efficiency will eventually commoditize the very services that keep these companies alive. As coding becomes cheaper, the strategic value shifts from "who can build it cheapest" to "who knows what to build." The winner isn't the firm with the best AI integration—it's the firm that controls the business logic, the domain expertise, the client relationship. Software development becomes the delivery mechanism, not the product.
Takeaway
The 25% number is a floor, not a ceiling. As AI capabilities compound, the cost reduction curve will steepen. The real question for investors and clients isn't whether IT services costs will fall—they will—but who captures the value differential.
Watch three signals: pricing model shifts from time-and-materials to outcome-based contracts; hiring patterns revealing whether mid-tier firms are upgrading their talent mix or just shrinking; and whether the AI cost savings flow to margins or get negotiated away by clients.
The IT services industry is being restructured from "selling hours" to "selling outcomes." Solvency is not a metric; it is a moment of truth. And for firms that can't adapt their pricing architecture, that moment is coming sooner than the market prices in.