AI Disruption Cut Infosys and TCS 30%. Valuations Haven't Reset.

Infosys and Tata Consultancy Services each lost roughly 30% of their market value over the past 12 months. The sell-off began when executives at both firms acknowledged that generative AI was changing how clients buy technology services. Several large customers have asked for cost reductions tied to AI tools, compressing billing rates and shortening project cycles.

TCS reported its first quarterly revenue decline in three years for the period ended March 2024. Constant-currency revenue fell 1.5% sequentially. Infosys cut its revenue guidance for the fiscal year twice, ending at a range of 4% to 7% growth – well below the 13% to 15% it had projected a year earlier.

The market reacted by marking both stocks down. At current prices, TCS trades at 24 times trailing earnings. Infosys trades at 22 times. Those multiples are down from their 2022 peaks of 33 times and 30 times, respectively. Yet they remain above the 10-year averages of 21 times for TCS and 18 times for Infosys.

Valuations have not fully adjusted for the structural revenue risk, several analysts said. The core question is how much AI will compress the traditional IT services model. Coding, testing, and maintenance work that once required large teams can now be done by fewer engineers using AI assistants. That reduces headcount growth and limits margin expansion.

The next test comes when large outsourcing contracts come up for renewal. Clients will face a choice: pay less for the same service enabled by AI, or migrate to workflows built around AI-native tools. Both paths pressure pricing.

Some analysts argue the 30% haircut prices in a moderate disruption scenario but not a severe one. Their internal models assume AI will replace 10% to 15% of current coding work over the next two years. If the adoption rate accelerates toward 20%, earnings could fall another 15% to 20% from current consensus estimates. That would push forward P/E ratios back above the historical troughs seen during past downturns.

A faster revaluation would require either a clear catalyst – a major contract loss, a deeper guidance cut, or a big client publicly announcing an AI-led restructuring – or a prolonged period of earnings disappointment that gradually resets expectations. Neither has materialised yet.

For now, the 30% decline has made the stocks cheaper but not cheap. The multiples still embed an assumption that revenue growth will stabilise and margins will hold. If AI adoption continues to accelerate, that assumption will break. The contract renewal cycle over the next six to 12 months will provide the first real data point.