Everyone is debating whether AI will replace jobs. The wrong comparison is “AI vs a ₹3 LPA employee.”
The real comparison in large service-based companies is AI versus thousands of currently underutilized employees.
Most people outside the IT services industry don’t realize how the model works. Large companies hire aggressively based on projected demand, future contracts, and growth targets. As a result, a significant percentage of the workforce can spend months on the bench between projects.
During this period, companies still pay salaries and bear infrastructure, office space, management overhead, training, insurance, hardware, and other operational expenses.
So when leaders talk about AI reducing hiring, they aren’t necessarily saying an AI agent is cheaper than a productive software engineer. They’re saying AI can handle a substantial portion of the work that would otherwise require maintaining a much larger workforce buffer.
If a company traditionally needed 100 engineers for active projects, 30-40 on the bench for future demand, and additional teams for support, documentation, testing, reporting, and repetitive operations, AI can significantly reduce that excess capacity.
An AI agent doesn’t take leave, wait for project allocation, or spend months on the bench. It scales instantly when demand spikes and down when demand falls.
This is why executives are talking about slower hiring growth—not because AI has achieved AGI, but because it is becoming good enough to automate a large percentage of repetitive and process-driven work.
The uncomfortable truth is that the first jobs affected won’t necessarily be the highest-skilled ones. The biggest impact will be on highly standardized, repeatable, and easy-to-measure work.
Instead of arguing whether AI is “better” than humans, the more important question is: What skills remain valuable when AI can generate code, write documentation, create test cases, analyze logs, answer support tickets, and execute workflows?
The answer is the same: system design, business understanding, problem-solving, product thinking, communication, leadership, and owning outcomes instead of tasks.
AI may not replace every engineer, but engineers who effectively leverage AI will likely replace teams that don’t.
The discussion is no longer about whether AI is coming. It is about how quickly organizations can restructure around it.
🚨 "TCS will not be hiring the kind of numbers it used to hire; AI agents may soon match TCS's employee count."
- Chairman N. Chandrasekaran.