In a significant forecast regarding the evolution of the Indian IT sector, Infosys chief executive and managing director Salil Parekh stated on Tuesday, October 13, 2026, that artificial intelligence (AI) services could account for nearly half of the company’s work within two to three years. Speaking at the Bloomberg New Economy Forum 2026, Parekh highlighted that while AI-related projects currently represent a smaller portion of the firm’s total operations, they are growing at a double-digit rate sequentially.

The core of this shift involves developing AI agents, utilizing coding tools, and modernizing legacy technology for various clients. According to Parekh, these services contributed more than 8 percent of Infosys’ revenue in the previous quarter. Based on current growth trajectories, Parekh estimated that these services could reach a share of 20 to 30 percent of the company’s work within just a few quarters.

The Rapid Expansion of AI Services at Infosys

The rapid growth of AI within the organization is not merely a theoretical projection but is rooted in current market demand. Parekh noted that the "huge demand" for AI-led projects provides a positive outlook for both Infosys and the broader technology services industry in India. For a large-scale development company like Infosys, the transition involves moving beyond simple integration to complex implementation and change management for enterprise clients.

While the current volume of AI work is still relatively small compared to the total scope of Infosys’ operations, the pace of adoption is accelerating. Parekh suggested that if current trends continue, the shift toward AI-centric projects will redefine the company’s service portfolio. This transition involves a wide range of technical tasks, from creating specialized AI agents to upgrading outdated systems into modern infrastructures.

For clients, this means that an app development company or a large IT integrator like Infosys is increasingly focusing on how to make AI functional and cost-effective. As businesses move away from experimental phases, the focus shifts toward achieving tangible outcomes through modern AI. This evolution is expected to influence how much companies spend on running these systems and which specific models they choose to adopt.

Strategic Workforce Training and Recruitment

To prepare for this shift, Infosys is heavily investing in its human capital. The company has a workforce of approximately 300,000 employees. To meet the demands of the AI transition, Parekh announced that the company hired 20,000 college graduates last year and plans to recruit another 20,000 graduates during the current financial year.

A critical component of this strategy is specialized training. New hires are trained in AI before they begin working with clients on implementation tasks. This proactive approach ensures that the workforce is equipped to handle the complexities of modern AI integration. Currently, approximately 80,000 employees at Infosys are already utilizing coding tools for both internal initiatives and client-facing projects.

This massive investment in human capital ensures that the company can serve as a reliable partner for clients who need to navigate the complexities of the AI landscape. By training staff specifically in AI before they reach the client-facing stage, Infosys aims to streamline the transition from traditional IT services to AI-led solutions.

Navigating Costs and Model Selection in the AI Market

As adoption of AI services expands among enterprise customers, the factor of cost has become a primary consideration. Parekh noted that clients are actively investigating whether they can achieve the same results using cheaper tokens and open-weight models compared to the most expensive, high-end models currently available.

This shift in client behavior creates a unique opportunity for IT service providers. Instead of just building a specific AI application, the role of the provider expands to include helping customers select, integrate, and manage various models at an acceptable cost. This involves navigating the different capabilities of various models to find the most efficient path for a specific business need, whether that be managing a social media account through automated tools or optimizing a large-scale logistics network.

The choice between high-cost, high-performance models and more economical options will likely dictate the volume of infrastructure required. For instance, if a client can achieve their goals using a smaller, more efficient model, the demand for massive computing power might stabilize. This dynamic forces IT companies to become experts in "right-sizing" AI for their clients’ specific budgets and goals.

Risks in Infrastructure and Data Sovereignty

Despite the optimistic outlook for AI service growth, Parekh warned that the physical and regulatory infrastructure supporting this growth faces significant challenges. Data has become central to the AI ecosystem, but it is not a simple commodity to move or store freely. Stricter requirements regarding data sovereignty and cloud infrastructure are currently limiting some capacity expansion.

Furthermore, the massive investments being made in data centers face scrutiny. Parekh pointed out that chips are "massively expensive" and are subject to intense competition. There is an ongoing question regarding how long these heavy investments in physical infrastructure will remain useful as technology evolves. He suggested that advances in foundation models could significantly reduce the cost of running AI systems, potentially changing the amount of computing capacity that businesses need to maintain.

[IMAGE_PROMPT: A conceptual image of a massive data center and server farm located in a high-tech facility. The architecture is sleek and modern. Digital "data" streams are visualized as glowing lines of light flowing between servers. Cinematic lighting, sharp details, photorealistic, 16:9 aspect ratio.]

The potential for new models to make inference several times cheaper could shift the market’s reliance on massive hardware clusters. Parekh noted that these effects might become more visible in a five-to-ten-year timeframe. Additionally, he cautioned about a possible backlash against the expansion of AI infrastructure, particularly in Western markets, where public sentiment regarding the technology may be out of sync with the economic forces driving investment.

The Future Landscape for Indian IT Services

The transition toward AI is viewed by Parekh as a significant opportunity for the Indian IT services industry. Rather than viewing AI as a threat that might undermine traditional roles, he described the current demand as "good news" for the sector. The integration of AI into the core of service offerings allows companies to provide more value in areas like coding, automated agent creation, and the modernization of legacy systems.

As firms move forward, they must balance the rapid pace of technological advancement with the practicalities of cost and infrastructure. For a large enterprise, managing a bank account or a media account often requires high levels of security and reliability; AI tools must be integrated into these workflows in a way that maintains those standards while reducing manual overhead.

The evolution of the industry will likely depend on how well these companies can navigate the "middle ground" of AI—finding the balance between cutting-edge capabilities and sustainable costs. By focusing on training their workforce and helping clients navigate the complexities of model selection, firms like Infosys are positioning themselves to lead in an era where AI is no longer a separate department but a core component of the service delivery model.

Conclusion

The projections made by Salil Parekh highlight a transformative period for Infosys and the broader IT landscape. By potentially moving half of its work toward AI-related services within three years, the company is signaling a major shift in how technology services are delivered. While challenges regarding chip costs, data sovereignty, and public sentiment remain, the immediate trajectory points toward a massive integration of AI into the core of enterprise operations.

For the reader, this means that the "AI revolution" is moving from the laboratory to the boardroom. Businesses will increasingly rely on IT partners not just to build software, but to navigate the complex ecosystem of AI models, costs, and infrastructure. As Infosys scales its training programs and continues to see double-digit growth in AI services, the integration of these technologies will become a standard component of modern business operations, impacting everything from internal coding processes to external client-facing tools.

Sources

• Infosys CEO says AI could account for half of company’s work within three years, asiantimes, 2026-10-14
• AI services may account for almost half of Infosys’ work in 2-3 years: CEO Salil Parekh, telecomlive, 2026-10-14