India’s infotech services industry finds itself in a moment of reckoning. The latest quarterly earnings from leading players tell a consistent story: of muted revenue growth and stubbornly flat margins accompanied by large deal wins. Clients are demanding up to 30-40% fee discounts, arguing that efficiency gains from AI should directly flow into lower delivery costs.
Most service providers have doubled down on AI adoption but are finding that it is yet to translate into value differentiation that can command premium pricing. The paradox is stark. The sector is working harder and transforming faster. Yet, it is unable to accelerate growth.
Also Read |
AI can power growth and still burn investors: Here’s how
Across boardrooms, the same questions are being asked: “Where is the value of AI? Why aren’t productivity gains showing up in company margins?” The problem isn’t technology; it’s the legacy work architecture within organizations. AI has been infused into existing delivery models without re-imagining the work itself. Middle managers are caught defending traditional roles and workflows remain process-bound instead of outcome-oriented, even as the workforce struggles to reskill fast enough.
Until firms redesign work from first principles by rethinking activities, roles, metrics, interactions and accountability between humans and AI agents, the gap between AI’s promise and its impact will continue to widen.
AI is rewriting work equations: EY Nasscom’s (AI)deation to Impact report examines what it takes to translate AI adoption into tangible business gain. This framework of analysis recognizes that AI’s impact is task-specific rather than role-wide. Every role is a combination of atomic tasks—some repetitive and easily automated, others judgement-intensive and ripe for augmentation, and still others that AI can amplify through data-driven insights.
Also Read |
Solow Paradox: The productivity puzzle of AI can be cracked
By applying three lenses of exposure (the extent to which any task lends itself to improvement in productivity through automation and AI), complementarity (the degree to which the task will require human oversight and therefore AI augmentation) and intensity (the frequency of tasks analysed in granular time units to estimate volume and effort), we assessed which tasks are most susceptible to AI infusion and where humans remain indispensable. This provides a more realistic estimate of where sustainable value can be captured and how future roles could be reconstructed.
This shift also redefines new role archetypes that reshape the workforce pyramid. Traditionally, the infotech sector scaled by hiring large entry-level cohorts, managed by supervisory middle layers and anchored by a thin apex of leadership. With AI absorbing transactional work, the base of this pyramid is hollowing out. Mid-levels are being reconstituted around cross-functional problem-solving, orchestration and innovation oversight. Various workforce shapes are starting to emerge.
Consider the case of a call centre. By deploying domain-specific bots alongside process-specialized agents, the bulk of routine queries were automated. Humans were kept in the loop only for exceptions or approvals requiring judgement. This redesign enabled the centre to handle the same call volumes with a nearly 80% lower headcount, freeing human talent to focus on higher-value issues. At the client level, such re-imagination translates to structural efficiency and new capacity creation.
The message is clear: the window for an incremental uplift has closed. AI adoption is not about sprinkling more tools across delivery, but about redesigning work from the ground up. Leaders who view it merely as a cost play will capture short-term savings at best. The true promise of AI lies in value maximization, opening up entirely new sources of value, creating business models, client solutions and competitive edges that did not exist before. Meaningful AI adoption, thus, is not about doing the same work cheaper, but about doing work in a fundamentally different manner.
Also Read |
The great AI shake-up: TCS layoffs hint of much worse to come for the IT sector
Decouple growth from headcount: Rethink the workforce equation where scale is no longer tethered to hiring curves. Growth must be measured in terms of business impact, not headcount, with commercial models shifting towards outcome-linked pricing, shared risk and faster time-to-value.
Redesign work from an AI-first lens: Break jobs into their atomic tasks to determine where AI can automate, augment or amplify. Re-imagine workflows to unlock efficiency, while balancing today’s delivery realities with tomorrow’s AI-first foundations.
Reinvent career paths for a skill-based future: Move away from linear, tenure-based progressions to fluid, lattice-shaped pathways. Skills, adaptability and impact, not tenure, will become the new currency of advancement, enabling talent to grow.
Use new tools to track human+AI team performance: Redefine how work is measured, governed and protected in hybrid teams. Establish new interaction models, guard-rails and risk protocols to ensure that human judgement and AI execution complement each other with transparency and accountability. The AI transition is a rare opportunity. But the gains will not appear by default. They must be designed with intent, grounded in organizational context and shaped by bold leadership.
The authors are, respectively, technology sector leader, and partner, people consulting, EY India.
