AI is putting pressure on the traditional IT-services business model

What happened?

Reuters reported on 9 October that TCS maintained a 24% operating margin in the September quarter despite increasing investment in AI. Its annualised AI revenue rose by nearly 20% quarter over quarter to approximately $3.1 billion.

Reuters also described the wider challenge facing India’s technology-services industry: AI can reduce the time needed to complete certain tasks, potentially putting pressure on traditional billing models based on the hours teams spend on a project.

At the same time, businesses are buying services that help them adopt AI, modernise applications and improve their operations. That creates a new source of demand even as older revenue models come under pressure.

Why does this matter?

Consider an illustrative example: a software team previously needed ten days to complete a task. AI tools help it finish the same task in six days.

If the service provider charges purely for the hours worked, it may earn less for that task. If it charges for the outcome delivered, it may have an opportunity to improve profitability by completing more valuable work.

But faster code generation does not automatically mean that every project becomes cheaper. Requirements, testing, security reviews, integration, maintenance and customer approvals can still take considerable time.

This is why AI can simultaneously create new opportunities and put pressure on established business models.

What should students learn from this?

Don’t measure your engineering ability only by how quickly you can produce code.

Developers need to understand requirements, identify edge cases, review AI-generated code, test applications, debug failures and explain how their work improves a business process.

AI tools may increase the amount of code one person can produce, but code volume alone does not establish quality or business value.

Practical exercise: Build a small application with an AI coding assistant. Record the time saved, then measure how much additional time is needed for testing, debugging and security checks. Document where the tool helped and where human judgement remained necessary.

The Aiexl. perspective

If AI can help you write code faster, what makes your work valuable beyond the code itself?

Cover: AI-generated editorial illustration, not a photograph of TCS staff.