At TCS’s thirty first annual basic meeting on June 9, N. Chandrasekaran, chairman of Tata Sons, described AI as the biggest progress alternative within the firm’s historical past. He predicted that TCS will quickly run as many AI brokers because it employs folks, and that governing AI methods will change into the following recurring income annuity for the agency.
The optimism is comprehensible, however the menace from AI brokers is extra speedy than the chance. The identical modernization work that constructed the Indian IT companies business is now being automated by the hyperscalers these corporations resell. Amazon’s AWS Transform deploys AI brokers that deal with the code evaluation, dependency mapping, refactoring and check era that TCS, Infosys, Wipro, HCLTech, Tech Mahindra and Cognizant have billed by the hour for 20 years. Extra importantly, one of many largest cloud platforms is providing the core modernization functionality at no further cost.
How Indian IT Constructed Its Annuity
To know the menace, we have to take a look at the evolution of the Indian IT companies mannequin. It started with labor arbitrage, as Indian corporations delivered software program improvement and upkeep at a decrease price than onshore groups. Customized software improvement adopted, as purchasers outsourced the constructing of their methods. The construct then hardened into software upkeep, the place one-time tasks grew to become long-running help contracts that billed yr after yr. Cloud migration and modernization packages opened the following income pool, and managed companies prolonged it. AI brokers now go after the repetitive work inside upkeep and modernization, the place the annuity, not the construct, was all the time the worthwhile half.
Inside AWS Remodel
At its core, AWS Remodel is an agentic modernization platform designed to automate the repetitive phases of software migration. It runs specialised brokers throughout three tracks: .NET, mainframe and VMware. The platform begins by discovering the appliance property, then analyzes the supply code and dependencies, generates a migration plan, refactors the appliance into a contemporary structure, and produces the exams that validate practical parity with the unique system. The output is just not solely transformed code however a modernization workflow that removes a lot of the guide inspection, rewriting and testing that giant engineering groups as soon as carried out.
Enterprises can level AWS Remodel at a legacy property and reduce the invention, planning and refactoring effort that giant modernization packages often demand. With AWS Remodel, Amazon is enabling enterprises to compress migrations that after ran for years into months, and it reviews clients have pushed greater than 1.1 billion traces of code by the service and saved over 810,000 hours of guide effort. For .NET workloads, the brokers run the conversion as much as 4 occasions quicker whereas eradicating the Home windows licenses that carried a lot of the working price. CIOs can now ask service suppliers to cost modernization on outcomes relatively than headcount, and platform groups can hand the brokers the repetitive evaluation and conversion as soon as performed by offshore supply groups.
The market leaders are publishing the identical low cost. Infosys informed buyers it moved 3 million traces of Hertz COBOL right into a microservices surroundings utilizing AI basis fashions at 60% decrease price and on a timeline 60% shorter than a traditional migration. A vendor that advertises an consequence delivered for 60% much less is instructing each future purchaser what that consequence ought to now price.
AWS Remodel vs The System Integrators
Indian IT majors are usually not standing nonetheless. TCS, Infosys and Wipro have constructed platforms reminiscent of MasterCraft, Topaz and Wipro Intelligence to automate components of the modernization lifecycle, and these belongings nonetheless matter the place area information, regulatory context and client-specific structure determine the end result. The problem is that these instruments are connected to a companies enterprise mannequin, whereas AWS Remodel is connected to cloud consumption. Amazon can afford to make the core service free as a result of it monetizes the vacation spot, the AWS infrastructure the place the modernized workload ultimately lands. With this, AWS is popping modernization right into a cloud acquisition movement relatively than a services-led transformation program. The important thing distinction lies in monetization, the place hyperscalers monetize the vacation spot whereas Indian IT corporations monetize the journey of getting there. Microsoft and Google Cloud are constructing the identical type of agentic modernization tooling, which makes this the path of the platform layer relatively than one firm’s wager.
Chandrasekaran’s strongest level is that context and belief change into the scarce assets, and that governing AI methods will want steady supervision. He’s proper that agentic companies like AWS Remodel produce plans and refactored code, whereas a group nonetheless has to deal with the fraction it can’t deal with, verify parity, handle the change, and join the output to how the enterprise runs. That supervision is unlikely to change into an annuity on the previous scale and margin as a result of governing AI runs on the hyperscalers, the mannequin labs, and Microsoft inside Azure AI, all of whom sit nearer to it than any integrator and intend to promote it themselves.
What The Shift Means For Every Facet
What does this imply to enterprise clients? How ought to CIOs take a look at modernization contracts within the age of brokers? For enterprise patrons, the shift strikes modernization from effort-based pricing towards outcome-based pricing and provides them a stronger negotiating place. For the Indian IT corporations, the previous annuity is shrinking, and the defensible place lies in governance, domain-specific brokers, business workflows and controlled AI operations relatively than in reselling the brokers the hyperscalers constructed. For the hyperscalers, agentic modernization turns into a buyer acquisition technique that pulls workloads onto their very own infrastructure.
The place The Strain Is Displaying
The strain already reveals in hiring, the place the 5 largest Indian IT corporations reduce a mixed 6,981 jobs in FY26, reversing a internet addition of 12,718 the yr earlier than. Cognizant has signaled plans to remove 12,000 to 15,000 roles in an AI-led restructuring that falls hardest on its India workforce.
Income is underneath strain as nicely, with Wipro’s IT companies revenue flat in fixed foreign money for the yr whilst massive deal bookings rose, and Infosys guiding FY27 progress to between 1.5% and three.5% whereas naming AI productiveness compression as a pressure working in opposition to it. Analysts mannequin 2% to three% annual deflation throughout the appliance companies base as brokers change the pricing mannequin for software work.
The visa surroundings provides one other layer, with H-1B approvals for the six greatest corporations down about 40% yr over yr, elevating the price of inserting Indian engineers in entrance of US purchasers on the identical time that automation reduces what number of are wanted.
The place This Leaves Indian IT
Enterprise clients are more likely to acquire leverage from this transition. A vendor whose historic revenue got here from the upkeep now being eliminated has little cause to modernize at full velocity. Enterprises can use that to cost these contracts on outcomes relatively than hours, and ask what the seller owns within the governance layer {that a} hyperscaler can’t replicate. The substitute alternative is real, however it’s neither captive nor confirmed on the identical scale as the upkeep annuity it goals to interchange.
The street forward for Indian IT corporations is determined by whether or not they can construct defensible governance, area and operating-layer IP round AI brokers relatively than reselling the brokers the hyperscalers ship. If they will, AI will increase their market relatively than erode it, and till then the safer studying is that brokers transfer the revenue towards the businesses that offer the intelligence and away from those which have lengthy built-in it.
