For India’s IT and business-process industry, the AI debate has moved beyond hype to hard economics: If automation begins to squeeze the labour-based revenue model that powered its global rise, what happens to pricing, margins and hiring? Jasjit Kang, managing partner and global head of business process services at Wipro—one of the sector’s largest operators—addresses those questions in an exclusive interview with Forbes India.
Kang explains that roughly 15 to 20 percent of entry-level revenue is structurally at risk and traditional headcount-based billing is giving way to outcome-linked contracts where providers commit to results and carry performance risk. He points out that more than 70 percent of his business is already AI-embedded, and he argues the shift from labour arbitrage to “knowledge and intelligence arbitrage” could expand the addressable market even as parts of the pyramid shrink.
For executives, investors and policymakers, the implications are immediate: Pricing models, workforce structures and competitive moats are all being renegotiated. Kang discusses what is genuinely at risk, what is over-stated, and how quickly the reset will show up in revenue, margins and valuations. Edited excerpts:
Q. What’s the industry mood? Are people optimistic about AI or concerned?
I think people who understand AI are optimistic. People who don’t, who are AI illiterate, are concerned. This is a different shift compared to what we’ve seen in the past. You’re putting the power of technology in the hands of the user. The operator who understands the domain can now design and implement. It’s a much bigger shift. But there’s an equally big opportunity.
If you break down AI, there are five layers—infrastructure, data, models, applications and orchestration. Whoever owns the process and data has a huge role to play. We own the process for most companies.
Q. Is AI a growth engine for Wipro’s business process services (BPS) division or a structural threat to the model India built?
It’s a growth opportunity very clearly. The pivot is happening from labour arbitrage to knowledge arbitrage and intelligence arbitrage. Labour arbitrage is something we left behind many years ago. The question now is: Where does the knowledge sit? People who understand the process understand the data. That’s the opportunity.
Q. How does the economics of business process services change? Is pricing moving from FTE (full-time equivalent) billing to outcome-based models?
Absolutely. You will see gain-share and outcome-based models. My personal view is that there will be some level of cannibalisation in the system to get to new models. But the value of work we deliver is going to go up significantly.
The way we measure RoI (return on investment) will also change. Are you looking purely at productivity, or are you looking at outcomes and value? That’s the shift.
We are moving away from pure cost arbitrage. Now think about a model where you layer skill, intelligence and knowledge with agentic AI. You create hybrid delivery models. As a result, commercial models will also start to change.
We will see more outcome-based and gain-share structures rather than traditional time-and-materials billing. New models will emerge—some purely agentic, some hybrid. If I had to bet, I would bet on hybrid in the short- to medium-term.
Q. In outcome-based AI contracts, who carries the model risk if something fails?
We underwrite the risk. We commit to an outcome, and we deliver it. That part does not change. What changes is that enabling productivity becomes easier.
Q. If generative AI automates 30 to 35 percent of repetitive work, why shouldn’t investors expect permanent margin compression?
That’s a wrong way of looking at it. Entry-level commodity jobs are going to get impacted. But knowledge work will increase. The ability of someone who understands the domain to architect workflows is much higher with AI. When you move one stage up the value chain, you typically see about a 15 to 20 percent uplift in value delivery.
Q. When productivity jumps 30 to 40 percent, who keeps the margin?
It depends on who invests. If the service provider is investing in building the capability, they will first recover that investment and then move into a gain-share model with the client. It could be 50-50, 60-40—it depends on the arrangement.
Q. What percentage of traditional services revenue is structurally at risk?
Entry-level jobs represent roughly 15 to 20 percent of revenue today. Those are at risk because they will get replaced. But when you remove one layer, you have investment available to reskill and move people up. Vendor consolidation will also happen. If you’ve done a good job, clients give you more work and reduce the number of providers.
Q. What percentage of Wipro’s BPS revenues today are AI-embedded and where does that number need to be in three years for the market to rerate the business?
Upwards of 70 percent today. It should be 100 percent. We’ll get there. A lot depends on client readiness. Clients need education and change management. They need to be comfortable with data and security. That’s where the delay is.
Q. If hyperscalers and model companies automate workflows themselves, what prevents disintermediation?
All the hyperscalers are working with us today. They have a niche. We have a niche. Overlaps will increase over time. But orchestration happens in the process. Any enterprise implementing AI needs people who understand the workflow and can bring the right AI solution. There are many AI solutions—someone has to integrate and run them.
Q. How much of your current BPS workforce is AI-augmentable versus AI-replaceable?
I would say 100 percent of the workforce is AI-augmentable. AI-replaceable… I’d put about 10 to 15 percent in the next 12 to 18 months. In customer experience, maybe 20 to 30 percent. Horizontals like HR operations, order-to-cash, procure-to-pay, finance and accounting will see some impact. But every time we replace a layer, we move that layer up the value chain.
Q. Is the future workforce smaller but more specialised, or larger and AI-enabled?
Larger and AI-enabled. You have to look at the entire value stream. With AI, you can move into new areas. That’s why the opportunity is big.
Q. What roles and skills will grow fastest, and which will shrink?
Traditional commodity roles will shrink. That’s inevitable. What will grow are skills around working with AI—people who are comfortable using it, building with it, understanding large language models, prompt engineering and so on.
You will see decline in entry-level repetitive roles because AI will automate a lot of that work. At the same time, new roles will emerge. You may see new talent models—for example, hiring specialised experts or PhDs for specific AI training work on a contract basis.
Organisational structures will also change. The pyramid will flatten. What the new structure looks like—whether it’s a diamond or something else—will evolve over time.
Q. If you rebuilt the business from scratch in an AI-native world, what would you do differently?
Earlier we built the model heavily dependent on people. The new model would be 50 percent technology and 50 percent people. You wouldn’t scale purely with human talent.
Q. Five years from now, will we still call this business process services? Will services start to look like products?
It could evolve—knowledge process services, for example. The core remains delivering value. But the shift is toward delivering outcomes. In a way, you will have service-as-software. You need to productise quickly.
Q. What proof points do investors need to see for the market to re-rate large IT and BPS firms?
Investors need to see integrated capability, the overall capability of an organisation. Do you have technology, data, domain and services together? If you’re strong in only one area, you’re at risk. Customers increasingly want a single partner who can deliver across the ecosystem.
Q. Is AI more deflationary for the industry or more expansionary than investors realise?
Strongly expansionary. Translation and automation will shift work globally. Sovereign AI initiatives will need scale and process knowledge. Where will that come from? India. I think overall it’s a big opportunity.
Q. When will the impact become visible?
We’ll start to see meaningful impact in the next 12 to 18 months. The technology is ready. Client readiness is the challenge. Once clients move, adoption will accelerate quickly.
First Published: Feb 26, 2026, 17:35
Subscribe Now