Infosys faces the AI test as coding automation rewrites Indian IT services

At Infosys’ 45th Annual General Meeting, Artificial Intelligence was not a passing reference. It was the question behind almost every serious question. The company had a steady year to report. Revenue crossed USD 20 billion in fiscal 2026. Constant currency growth stood at 3.1%. Adjusted operating margin was 21%. Free cash flow was strong at USD 3.7 billion. Large deal total contract value touched USD 14.9 billion, with 55% of it coming from net new business. On paper, these are not numbers of a company in crisis. Yet, the mood around the AGM showed something deeper. Shareholders were not only asking how Infosys had performed. They were asking what happens to Infosys when Artificial Intelligence begins to automate coding, change delivery models, reshape talent needs, and alter the economics of technology services. That made the AGM more than a shareholder meeting, it in a way served as platform to look at the evolving trajectory of not just Infosys, but overall IT Services industry. The existential question Nandan Nilekani, Non-Executive Chairman, put the issue on the table directly. Every major technology transition, he said, brings questions about relevance, leadership, growth, and margins. With AI, those questions have become pronounced because the shift is larger and more disruptive. The hardest question is also the simplest one. If coding becomes automated, why are IT services companies needed at all? Infosys’ answer was clear. Coding is important, but enterprise technology is not just coding. Large organisations need systems that work with existing investments. They need testing, resilient architecture, cybersecurity, data governance, compliance, and deep knowledge of how their business actually runs. AI tools may write or improve code, but they do not automatically understand a bank’s risk systems, a retailer’s supply chain, a manufacturer’s shop floor, or a telecom company’s customer operations. This is where Infosys wants to position itself. It is not trying to tell investors that AI will leave the services model untouched. It is saying the model will change, and that the companies which can adapt will become more relevant. The phrase that mattered most The most important insight from the AGM was “AI deployment gap.” That gap is where Infosys sees its next opportunity. Many enterprises have experimented with generative AI. Fewer have managed to turn it into production-grade business impact. The challenge is not only model access. It is execution. It is the hard work of putting AI inside core systems, connecting it to enterprise data, securing it, governing it, and making it useful to people who run real business processes every day. This is a very different story from the early excitement around chatbots and copilots. Infosys is seeing enterprise AI as an integration and transformation problem. In that view, the AI success will not be decided only by who the models. They will also be decided by who can make models, agents, data, cloud, cybersecurity, and transaction systems work together and make for seamless delivery of outcomes. For Indian IT services, this is the most important part. If AI is seen only as coding automation, it can look like a threat. If AI is seen as enterprise deployment at scale, it becomes a large services opportunity. There is no use in the use cases. We need to real, on ground projects and how AI has been intersected with those projects and how it changed the outcomes. Why legacy modernisation is back One striking point from the AGM was Infosys’ view that AI has made legacy modernisation urgent in a way nothing else has. For years, modernisation was often sold as a cloud migration or cost optimisation exercise. AI changes that conversation. Enterprises cannot build serious AI-led workflows on top of fragmented data, ageing applications, brittle architecture, and disconnected systems. If they want agents to act on business processes, the underlying technology estate has to be cleaner, more connected, and more secure. That gives Infosys a familiar but renewed market to address. Modernisation is no longer only about moving old systems to the cloud. It is about making the enterprise AI-ready. This also explains why Infosys continues to talk about Topaz, Fabric, and Cobalt together. Topaz and Fabric represent the AI layer. Cobalt brings the cloud foundation. The larger message is that AI cannot scale in isolation. It needs cloud, data, engineering, governance, and domain knowledge around it. The six AI value pools Salil Parekh, Chief Executive Officer and Managing Director, described six broad areas where Infosys sees clients using AI. These are AI engineering and strategy, data, process transformation, technology modernisation using agents, physical AI, and AI trust. This framework in a way gives a direction and it shows that Infosys is trying to move the conversation beyond “AI can write code.” The company is presenting AI as a full enterprise transformation stack. For instance, AI engineering and strategy is about building agents and AI-led applications. Data is about making structured and unstructured enterprise data usable. Process transformation is about improving workflows before placing agents inside them. Technology modernisation is about using AI to renew ageing systems. Physical AI takes the story into manufacturing, automotive, medical devices, and other product environments. AI trust focuses on security, responsible AI, and governance. When one connects all the dots, Infosys sees these areas as part of a USD 300 billion to USD 400 billion AI-first services opportunity by 2030. That is the big market-sizing claim. The real question however, will be how much of this opportunity turns into sustained revenue growth, better margins, and stronger market confidence. Data is the real battlefield One of the strongest points that emerged from the AGM was the importance of data. Infosys’ argument is that enterprises will not get lasting value from generic AI alone. The value will come when their own data is made available to AI models in a secure, governed, and useful way. This is where the services opportunity becomes clearer. Many enterprises still have data spread across applications, geographies, business units, cloud environments, and legacy systems. Some of it is structured. Much of it is unstructured. Some of it is clean. Much of it is not. Without fixing that foundation, AI may remain stuck in pilots. For Infosys, this creates work across data engineering, cloud migration, application modernisation, cybersecurity, compliance, and business process redesign. In other words, AI is becoming the front door through which older enterprise technology problems are being reopened. This is also why Indian IT services firms may still have a role in the AI era. The problem is not only to build an agent. The harder problem is to make that agent work safely and usefully inside a live enterprise. AI revenue is now under scrutiny The AGM also showed that investors want more than broad AI messaging. They want numbers. Infosys has disclosed that AI services contributed about 5.5% of revenue, or roughly USD 1 billion annualised. In the context of an IT services company, this refers to client work where AI is a core part of the consulting, engineering, data, modernisation, process, or trust layer, rather than revenue from selling a standalone AI model. The company also said this revenue was growing faster than the company average. It said it is working with 90% of its top 200 clients on their AI journeys, and has thousands of AI projects underway. These numbers are to be read with cautious optimism because they also raise the next set of questions. How much of this AI revenue is truly incremental? How much is replacing older work? Are AI services better for margins or do they carry investment pressure? Will productivity gains reduce billable effort? Will clients expect savings to be passed back? Can AI-led deals become large enough to offset softness in traditional discretionary spending? These were not abstract concerns. Shareholders asked about AI revenue, AI margins, separate AI reporting, growth, and the impact of automation on the services model. The market is clearly moving from excitement to accountability. The workforce reset The other big question was people.For decades, Indian IT services grew on the strength of talent scale. AI now challenges parts of that model. If tools can write code, test software, generate documentation, improve productivity, and support agents, the old relationship between headcount and revenue will change. Infosys did not present this as a job-loss story. Its line was about reskilling, redeployment, and new kinds of work. The company said it hired more than 20,000 college graduates during the year and ended with a workforce of over 325,000 employees. It also spoke about preparing employees for the AI era, and redeploying people released through productivity gains into new growth areas. That data is important. It suggests that the work mix is changing, even if the company is not saying it as workforce reduction. The future model may not be only about how many people are added. It may be about how effectively employees, domain experts, platforms, and agents work together. This is where Infosys’ internal skilling becomes critical. The company spoke about employees being AI aware, AI builders, and AI masters. That progression matters because awareness alone will not be enough. The next phase will need people who can design, deploy, govern, and improve AI-led systems inside complex enterprise environments. Shareholder anxiety was visible The AGM also revealed an emotional undercurrent. Many shareholders praised Infosys for governance, dividends, CSR, leadership, and transparency. But the repeated concerns were hard to miss. They asked about the falling share price, buyback, bonus shares, AI’s impact on the IT services model, employee retention, AI revenue, and future growth. This is the real pulse of the company’s AGM. Infosys continues to command trust among long-term shareholders, but that trust now comes with a demand for proof. The question is no longer whether Infosys understands AI. The question is whether Infosys can show that AI will create visible and tangible business momentum. Investors want to know if AI will lift growth, protect margins, support talent transformation, and make the company more competitive in a slow-growth market. The company’s response was measured. It declined to comment on share price movement. It pointed to its capital allocation policy, under which it expects to return around 85% of free cash flow cumulatively over a five-year period through dividends, buybacks, or special dividends. It also said its mergers and acquisitions approach would remain disciplined, focused on tuck-in acquisitions that fill capability gaps or strengthen growth areas. But the shareholder mood suggests that capital return alone may not be enough. Investors want Infosys to invest in the future, explain the AI revenue path more clearly, and show that it can convert AI capability into growth. Execution is the real risk Nilekani’s most candid clarification came when he explained execution risk. The opportunity, he said, is large because AI makes many new things possible. But Infosys has to reorient its services to AI-first and AI-augmented models. It also has to align sales and delivery, transform talent, and look at new pricing models, including outcome-based pricing. That is the crux of the matter. Infosys has the relationships, scale, balance sheet, delivery depth, cloud capability, and AI partnerships. But AI is forcing every large IT services firm to rethink the old playbook. Productivity gains can help margins, but they can also reduce revenue if clients demand the benefit. Outcome-based pricing can create upside, but it also transfers more risk to the provider. Agents can make delivery faster, but they also need governance, security, monitoring, and accountability. This is why the AGM felt important. It showed that the AI story has moved beyond pilots It is now about execution. Up ahead Infosys’ guidance for fiscal 2027 remains cautious, with revenue growth of 1.5% to 3.5% and operating margin of 20% to 22%. That caution reflects the broader environment. Client spending remains selective, and discretionary technology budgets are still under pressure. That makes the AI opportunity even more important. Infosys needs AI to become more than a positioning theme. It has to become a growth engine, a productivity lever, a client relevance marker, and a talent transformation programme. The AGM showed that Indian IT’s AI transition is already here. It is visible in shareholder questions, client strategy, workforce planning, delivery models, and revenue conversations. Infosys believes the puck is coming to where it has positioned itself. The market will now ask whether it can move fast enough. For Infosys, the AI challenge is not about proving that AI matters. That argument is over. The real challenge is proving that AI can translate into sustained growth, stronger margins, future-ready talent, and renewed confidence in the Indian IT services model.