Ravi Chauhan, CEO of Hexaware Technologies, told the press that AI could reduce IT project costs by 25%. The market treated this as a positive signal. I treat it as a defensive confession. That number is not an efficiency metric. It is an admission that the industry's pricing power has shifted to the buyer. And when a mid-tier IT services firm starts volunteering discount figures, it means the negotiation is already over.
The statement appeared across business media with minimal scrutiny. Crypto Briefing carried it as a standalone item. No data. No methodology. No breakdown of which delivery phases would absorb the cut. Just a CEO's prediction wrapped in a consultant's confidence. The information density of the original report is so low that any serious analysis must work backwards from industry structures rather than forward from evidence.
The industry's cost structure makes a 20-30% reduction mathematically plausible. Labor represents 60-80% of project costs. AI agents can compress coding, testing, and documentation cycles. McKinsey and Gartner both estimate 20-40% efficiency gains across typical delivery pipelines. The 25% figure sits comfortably inside those bounds. But that is precisely the problem.
Plausibility is not proof. It is a starting point for skepticism.
The deeper issue is that the CEO had to say this at all. The IT services sector operates on a time-and-materials model. Clients buy hours. AI compresses hours. That math should be a threat, not a promise. Yet he framed the AI-driven cost reduction as a benefit. That framing hides the structural dilemma.
Consider the revenue side. If costs drop 25% but billing rates drop by the same amount to remain competitive, margins do not improve. Revenue just shrinks. The CEO mentions balancing productivity gains with pricing to preserve margins. That phrase is a tell. It reveals the sector's real vulnerability: clients now expect a discount because they know the tooling exists. The negotiation has moved from “can AI do the job?” to “how much will you reduce the bill?”
The CEO's statement is not a prediction. It is pre-emptive capitulation.
I have audited this kind of narrative before. Not in IT services, but in crypto. Whenever a protocol team starts “pre-announcing” the efficiency of a new feature or the cost savings of a migration, I check the actual data. In crypto, the token price does the talking. Here, the customer does.
The structural pressure is not limited to Hexaware. The global IT services market is a $1.2 trillion ecosystem. It employs roughly 5 million people in India alone. If AI truly delivers a 20-25% cost reduction across the industry, that represents between $240 billion and $300 billion in annual value. That value will flow somewhere. The question is who captures it: the vendor, the client, or the AI platform providers like OpenAI and Microsoft.
Vendors will not capture it all. The client has the leverage. The buyer knows the tooling. They have read the same McKinsey reports. They have seen the same TCO comparisons. The vendor no longer controls the information asymmetry that once guaranteed premium pricing.
Competition will now be defined by the AI leverage ratio. The firms that can deliver outcomes at a lower cost will win. The firms that still sell hours will lose. The market is moving from ‘‘how many developers do you have?’ to ‘‘how few can you use to finish this?’ That shift is the real story. The 25% figure is just the visible symptom.
Mid-tier players like Hexaware are in the most fragile position. The top-tier firms—Accenture, TCS, Infosys—have the balance sheet to invest in proprietary AI platforms. Accenture has committed $3 billion. Infosys built Topaz. Wipro put in $1 billion. A mid-size firm cannot match that. Their move is to make the AI narrative public, and to hope that the market reads it as innovation rather than necessity.
That's a communication strategy. It is not a technology advantage.
There is a role for the bulls here. The contrarian case is not that AI is irrelevant. It is that the adoption timeline is faster than the market expects. If AI tools are already delivering 15-20% efficiency gains in production, then the 25% claim is conservative. The client-side demand for AI-native delivery is rising. This could create a new, high-margin segment for vendors who can quickly move to outcome-based pricing. The firms that solve the ‘‘maintain margin while cutting costs’’ dilemma could actually become the next majors. The first mover in outcome-based pricing wins the premium. The laggards get left in the time-and-materials ghetto.
But here is the uncomfortable part. If that scenario plays out, the sector's workforce will not be spared. The industry is still the largest employer of entry-level programming talent in India. A 20-25% cost reduction implies a direct reduction in labor intensity. The IDC projects 15-20% of junior coding roles will be automated by 2028. That is a structural shock. The mid-tier firm that already relies on low-cost labor arbitrage is the most exposed.
From my audit experience, I have seen this pattern before. When a protocol announces “we’ll lower gas fees for users,” it is usually because they have to. When an IT services CEO says “AI will cut costs,” it is because the client is already asking for it. The market is the ultimate auditor. The news cycle is just a window.
The real question is not whether AI will cut IT costs by 25%. It is whether the vendor can keep its margin after the cut. The CEOs talk about efficiency. The CFOs know it’s about pricing. And the clients? They are just waiting for the other side to make the first move.
The next 12 months will be decisive. Watch for three signals. First, whether Hexaware or its peers launch a public AI platform, not just a narrative. Second, whether quarterly earnings show AI-related revenue lines or just cost-savings press releases. Third, whether the industry shifts to outcome-based contracts in the next two quarters. If the shift is slow, the 25% cost reduction is just a rumor. If the shift is fast, the old business model is dead.
Volatility is just liquidity leaving the room. And in the IT services market, the liquidity is leaving the human labor model. The question is who will catch it.
Trust is a variable I refuse to define. A CEO's quote is just a input. The output is the contract. The margin. The balance. The proof is not in the press release. It is in the next quarterly report.
Until then, the 25% is a promise. And in this market, promises are liabilities.