AI Is Rewriting India’s IT Business Model
How artificial intelligence is changing pricing, productivity, hiring and
the economics of India’s technology-services industry
By Murthy ASR | October
1, 2026
India's
technology-services industry is entering a period in which artificial
intelligence is changing not only the technology companies use, but also what
they sell, how they price it and how much human labour is required to deliver
it.
That transformation is
becoming increasingly visible as India's major IT companies prepare to report
their September-quarter results.
Reuters reports that
India's roughly $315 billion technology-services sector is expected to
post its weakest sequential performance in about three years in the
July–September quarter, with analysts expecting revenue growth of only about 0.7%
to 3.5% quarter-on-quarter for the six largest companies. TCS is scheduled
to begin the earnings season on October 8, followed by other major companies
later in the month. (Reuters)
The numbers, however,
tell only part of the story.
The deeper change is
occurring inside the business model that made Indian IT services a global
success.
For decades, India's
technology-services industry grew by combining large pools of skilled engineers
with lower labour costs, global delivery centres and contracts that often
linked revenue to the amount of work or number of people deployed.
AI is beginning to
challenge that formula.
From manpower to productivity
The traditional Indian IT
model was built around scale.
A global company might
outsource application development, testing, infrastructure management or
business processes to an Indian technology provider. The provider would
assemble a team, allocate engineers to the project and charge according to the
agreed commercial model.
The more efficiently the
provider could deliver the work, the greater the potential margin.
Generative AI and coding
tools change that equation.
If an engineer equipped
with AI can accomplish work that previously required several people, the client
has an obvious question:
Why should it continue
paying for the same amount of human effort?
That question is
increasingly influencing contract negotiations.
Reuters reported in
August that Indian IT companies were shifting more contracts toward performance
and measurable outcomes rather than simply hours worked. TCS, for example,
said its CEO had indicated that about 80% of its contracts in finance, human
resources and other business services were based on outcome-performance
measures. (MarketScreener)
This is an important
structural change.
The customer is
increasingly interested in the result—not necessarily the number of employees
required to produce it.
The “AI deflation” problem
This creates what
analysts have described as AI-led deflation.
Imagine a software
project that historically cost $10 million because it required a large team
working for several years.
If AI reduces the amount
of work required by 20%, the customer may reasonably expect some of that
productivity gain to appear in the price.
The technology company
therefore faces a difficult choice.
It can:
- keep the old price and potentially increase
its margin;
- pass some savings to the customer and protect
the relationship;
- invest the productivity gain in additional
services; or
- develop new AI-related offerings capable of
generating additional revenue.
The market increasingly
expects the fourth option to become important.
Kotak Institutional
Equities estimates that AI adoption could create gross revenue deflation of
around 7% and net deflation of about 3.5% in parts of the IT-services
business. It also says AI-related revenue opportunities are not yet sufficient
to completely offset the pressure on existing services. (The Financial Express)
This does not mean that
the IT industry is disappearing.
It means the economics
of delivering IT services are changing.
The employment question
Perhaps the most
sensitive consequence concerns employment.
India's technology
industry employs millions of people. NASSCOM's 2026 strategic review estimates
direct technology-sector employment at approximately 6 million in FY2026,
with net additions of around 135,000 employees. (NASSCOM Community)
But the composition of
those jobs is changing.
For many years, the
industry's famous pyramid structure depended on bringing large numbers of
graduates into entry-level technology roles. Employees could begin with
testing, maintenance, coding or support work and gradually move toward more
specialised responsibilities.
AI-assisted development
threatens to reduce the amount of routine work available at the bottom of that
pyramid.
That does not necessarily
mean fewer technology jobs overall.
It could instead mean a
different mix of skills.
NASSCOM says hiring is
expected to shift from volume toward skill mix, reflecting productivity
gains from AI. (NASSCOM Community)
In practical terms,
demand may increasingly favour people who can combine software engineering
with:
- AI implementation
- data engineering
- cybersecurity
- cloud architecture
- domain expertise
- AI governance
- product development
- business-process redesign
The challenge is that an
experienced AI engineer cannot necessarily be created overnight from a
traditional entry-level coding position.
That creates a transition
problem for the industry.
Productivity does not automatically mean job destruction
There is another side to
the employment debate.
AI productivity can allow
companies to take on work that was previously too expensive or too
time-consuming.
A company that needs
fewer engineers to maintain an existing application could potentially use those
engineers to build new products.
This is why the
relationship between AI and employment is not as simple as “AI replaces
workers.”
Wipro, for example, said
in September that its AI initiatives had created productivity capacity
equivalent to the work of about 20,000 employees, with those workers
redeployed within the company. (MarketScreener)
That is a significant
distinction.
Automation can eliminate
tasks without necessarily eliminating the people performing them.
But whether redeployment
remains possible at scale will depend on whether companies can generate enough
new work and whether employees have the skills required for it.
TCS: From outsourcing giant to AI-led technology company?
Tata Consultancy Services
provides one of the clearest examples of the transition.
TCS remains one of the
world's largest IT-services companies, but its own strategy increasingly
describes AI as a fundamental transformation rather than another service line.
The company says it
generated more than $30 billion in consolidated revenue in FY2026 and
has set an ambition to become the world's largest AI-led technology-services
company. (Tata Consultancy Services)
The significance of that
ambition is greater than the wording suggests.
An AI-led TCS cannot
simply depend on the old formula of adding more engineers as revenue grows.
It needs to move higher
up the value chain.
That means helping
clients redesign business processes, deploy AI systems, manage data and
infrastructure, and eventually operate increasingly autonomous enterprise
systems.
In other words:
The product is gradually
shifting from labour to intelligence.
The infrastructure opportunity
There is an intriguing
contradiction at the heart of the AI revolution.
AI can reduce the amount
of human labour required to perform certain software tasks.
But AI requires enormous
amounts of computing power.
That creates a completely
new infrastructure market.
TCS subsidiary HyperVault
has secured 264 acres in Hyderabad for a planned AI data-centre campus with
capacity of up to 1 gigawatt. TCS says the project could involve
investment of up to ₹70,000 crore and will be developed in phases for AI
companies and hyperscalers. (Tata Consultancy Services)
This illustrates an
important economic paradox.
AI may reduce the amount
of human labour required to produce software while dramatically increasing the
amount of computing infrastructure required to produce AI-powered services.
That creates
opportunities for businesses involved in:
- data centres
- GPUs and semiconductor infrastructure
- power generation
- electricity transmission
- cooling systems
- networking
- cloud computing
- cybersecurity
- construction and engineering
JLL estimates that
India's data-centre capacity could rise from around 1.6 GW in mid-2026 to 6
GW by 2029, requiring very substantial investment in infrastructure. (JLL)
The new AI infrastructure economy
Hyderabad's emerging role
The geographic
implications are particularly interesting for India.
Hyderabad has
increasingly positioned itself as a major technology and data-centre hub.
The planned HyperVault
facility adds to that momentum, while Microsoft's expansion of cloud
infrastructure in Hyderabad is another indication of the region's growing
importance in India's AI infrastructure ecosystem. (The Times of India)
This creates a second
layer to India's AI story.
The first layer is AI
services—companies helping global businesses use AI.
The second is AI
infrastructure—the physical computing capacity required to run those
systems.
The economic value could
therefore extend well beyond traditional software companies.
The rise of outcome-based contracts
Perhaps the most
important change for investors is happening inside contracts.
Under the old model, a
technology company could largely monetise human capacity.
Under the emerging model,
the customer increasingly wants to monetise the outcome.
Suppose a bank wants to
automate its customer-service operation.
Previously, the
technology provider might have supplied a team of hundreds of employees.
Under an AI-enabled
model, the provider may instead be paid according to:
- response times;
- automation rates;
- cost savings;
- resolution rates;
- customer satisfaction;
- system availability.
That fundamentally
changes the commercial relationship.
The technology company
assumes more responsibility for productivity.
The customer, in turn,
expects to share in the resulting savings.
Reuters reported that
such outcome-linked arrangements are becoming increasingly common across
India's IT-services industry. (MarketScreener)
This could eventually
make traditional measures such as employee count and revenue per employee
less useful on their own.
Investors may
increasingly have to examine metrics such as:
Revenue per AI-enabled
employee
AI-related revenue
Gross productivity
savings passed to customers
Recurring platform
revenue
Outcome-based contract
value
Proportion of contracts
linked to business outcomes
Why the September quarter matters
That brings the story
back to the current earnings season.
The September quarter is
important because investors will be listening for evidence of whether AI is
beginning to affect financial results in a meaningful way.
Reuters reports that
analysts expect weak sequential growth among the major IT companies, while
investors are likely to focus heavily on companies' full-year revenue guidance.
(Reuters)
The key questions will
include:
1. Are clients spending more on AI?
If yes, how much of that
spending is genuinely incremental?
2. Is AI revenue replacing traditional revenue?
A company can announce
large AI contracts while simultaneously losing revenue from traditional
services.
3. Who receives the productivity benefit?
If AI allows a project to
be completed with 30% fewer resources, does the client keep the saving, does
the IT company keep it, or is it divided?
4. Are contracts becoming shorter?
Reuters has reported that
uncertainty around AI is contributing to shorter contracts in some cases. (MarketScreener)
5. What happens to entry-level hiring?
This could become one of
the most important long-term indicators of the industry's transformation.
Infosys illustrates the other side of the equation
Infosys provides another
useful example.
Its FY2026 annual report
says AI-led programmes had been deployed across 90% of its top 200 clients,
while the company reported FY2026 revenue of ₹1,78,650 crore and 3,28,594
employees globally. (Infosys)
That combination is
revealing.
AI is simultaneously:
a threat to traditional
services
and
a major new source of
business.
The challenge for Indian
IT companies is therefore not simply to adopt AI.
They must adopt it faster
than the productivity gains reduce the value of the services they currently
sell.
That is the central
economic problem.
Wipro's transformation
Wipro is pursuing a
similar strategy.
In August, Wipro
announced an expanded partnership with Google Cloud aimed at scaling Gemini
Enterprise and agentic AI across enterprise operations. It said it would deploy
more than 10,000 AI-certified specialists, including 1,500
forward-deployed engineers. (Wipro)
At the same time, Wipro's
June-quarter results showed IT-services revenue declining 1.4% sequentially
in reported terms, while its large-deal bookings increased 12.9%
sequentially in constant-currency terms. The company guided for
September-quarter IT-services revenue between $2.574 billion and $2.627
billion, equivalent to a constant-currency sequential range of -1.5% to +0.5%.
(Wipro)
The combination captures
the industry's transition.
Deal activity can remain
healthy even while near-term revenue growth remains weak.
Why?
Because the nature of the
deals is changing.
A large AI transformation
contract may initially generate less revenue than the equivalent
labour-intensive outsourcing contract—but could create a different long-term
revenue model.
The smaller-company challenge
AI may also alter the
competitive hierarchy within Indian IT.
The old industry rewarded
scale.
A company with 200,000
engineers could deploy huge teams across hundreds of projects.
But if AI dramatically
increases individual productivity, the advantage of having a very large
workforce could diminish.
A smaller technology
company with highly specialised engineers could potentially compete for
contracts that once required a much larger workforce.
Reuters has reported that
mid-sized companies such as Persistent Systems and Coforge have recorded
stronger recent growth than several of the traditional giants, while clients
increasingly demand flexibility and productivity. (MarketScreener)
This does not mean
smaller firms will automatically replace the large companies.
Large IT providers retain
major advantages in global delivery, customer relationships, cybersecurity,
compliance, infrastructure and domain knowledge.
But AI may reduce one
historic advantage:
the sheer size of the
workforce.
What happens to India's IT advantage?
This is perhaps the most
important question for policymakers.
India's IT-services
success was built on a combination of:
talent + scale + cost
advantage + English-language capability + global delivery infrastructure.
AI potentially weakens
the importance of the scale-plus-cost component.
If one highly skilled
engineer using AI can perform the work previously handled by several engineers,
the cost difference between countries may become less important.
That could be a threat.
But it could also be an
opportunity.
India has a very large
technology workforce, a deep engineering talent pool, established relationships
with global companies and a growing digital infrastructure ecosystem.
If those advantages can
be combined with advanced AI capabilities, India could move from being
primarily a low-cost technology-services provider toward becoming a high-value
technology and AI partner.
NASSCOM describes this as
a shift from scale-led growth toward value and innovation. (NASSCOM Community)
A new definition of productivity
The biggest mistake would
be to measure AI's impact only by the number of jobs eliminated.
The more important
economic question is:
What does the economy do
with the productivity created by AI?
If AI allows an Indian IT
engineer to deliver three times as much output, there are several possible
outcomes.
The client could pay
less.
The IT company could earn
more.
The employee could handle
more sophisticated work.
The company could take on
additional customers.
Or some combination of
all four could happen.
The distribution of that
productivity gain will determine whether AI becomes primarily a margin
story, a growth story or a labour-market story.
Most likely, it will be
all three.
The next phase of India's IT industry
The Indian technology
industry is not entering a simple decline.
It is entering a
transition.
The evidence already
points in several directions.
Traditional IT-services
growth is under pressure. AI is creating pricing pressure. Clients are
demanding measurable outcomes. Hiring is becoming more skills-focused. At the
same time, companies are investing heavily in AI capabilities and
infrastructure. (Reuters)
And the infrastructure
investment is enormous.
TCS's planned 1GW
Hyderabad AI campus is one example of the capital being deployed behind the
next generation of computing. (Tata Consultancy Services)
That suggests India's
technology story may be moving from one dominated by people and software
toward one combining:
people + software + AI
models + computing infrastructure + electricity + data.
The bottom line
The defining question for
India's IT industry is no longer whether companies will adopt artificial
intelligence.
They already are.
The more difficult
question is who will capture the economic value created by it.
Will it go to global
customers through lower prices?
Will Indian IT companies
retain it through higher margins?
Will new AI products
create entirely new revenue streams?
Will employees become
dramatically more productive?
Or will the savings
ultimately be distributed among customers, companies and workers?
The September-quarter
earnings season will provide some early clues, but the transformation will take
years to complete.
For India's technology
industry, AI is therefore not simply another technology cycle.
It is a potential change
in the industry's underlying business model.
The companies that
succeed may not necessarily be those with the largest workforces.
They may be those that
can turn AI-driven productivity into sustainable revenue, higher-value
services and new forms of customer partnership.
That is the real business
story unfolding in India's IT sector.
Key numbers to watch
|
Indicator |
Current reference |
|
Indian technology sector |
~$315 billion |
|
Direct technology employment |
~6 million |
|
Expected Q2 sequential growth, top six IT firms |
0.7%–3.5% |
|
TCS planned Hyderabad AI campus |
Up to 1 GW |
|
TCS/HyperVault planned investment |
Up to ₹70,000 crore |
|
India's projected data-centre capacity by 2029 |
~6 GW |
|
Infosys FY2026 revenue |
₹1,78,650 crore |
|
Infosys global employees |
3,28,594 |
Sources: Reuters,
NASSCOM, TCS, Infosys and JLL. (Reuters)
Comments
Post a Comment