India’s technology sector is sitting on a structural fault line that quarterly earnings have not yet fully exposed. Motilal Oswal Financial Services, in its sector update dated May 4, warned that large-cap IT revenues will decelerate in FY27, as artificial intelligence-driven cost compression accelerates faster than new AI revenue opportunities can replace what is being lost.
AI deflation is now measurable
Motilal Oswal said in its May 4 report that the cost of delivering an IT services outcome has fallen sharply, and those savings are currently being extracted almost entirely by enterprise buyers rather than flowing back to service providers or model companies.
The brokerage used a direct illustration from Sayandeb Banerjee, co-founder and chief executive of MathCo a Bengaluru and Chicago-headquartered enterprise AI company with roughly Rs 670 crore in annual revenue to bring forth the basis of their assumption.
According to this, cloud migration previously required a 500-person team working for 18 months on time-and-materials billing. The same migration exercise under AI compresses to roughly 50 people, meaning revenue per workload falls even when total workloads across the economy grow.
“There is genuine real-term revenue compression. The cost of delivering an outcome has fallen sharply; these savings are currently being extracted almost entirely by enterprise buyers through productivity demands and vendor price pressure,” Motilal Oswal quoted Banerjee as saying in the report.
The brokerage also noted that Google, Anthropic, and OpenAI are not yet the economic winners from this compression, as they are absorbing heavy infrastructure costs and actively subsidising enterprise adoption. A hyperscaler may commit $100 million in services spend alongside an enterprise Gemini rollout just to ensure deployment succeeds, the report cited as one example.
As token costs normalise and subsidisation reduces, the value pool will need to be distributed across enterprises, model providers, and service ecosystem participants a redistribution that current IT services valuations may not be adequately pricing in, Motilal Oswal said.
The context moat is real but shrinking faster than expected
Motilal Oswal Financial Services Ltd. acknowledged that deep knowledge of enterprise systems, processes, and data remains the central variable in enterprise AI return on investment and that system integrators and business process management firms have genuine defensibility built over decades.
But the brokerage said AI is democratising context acquisition at a pace that even seasoned practitioners did not anticipate.The brokerage said the AI-as-divider dynamic means even the implementation multiplier assumption embedded in the bull case for the sector needs to be revisited.Motilal Oswal’s prior analysis, referenced in the May 4 note, estimated that 12% to 15% of sector revenue faces direct AI-driven displacement.
Why outcome pricing claims from large IT firms deserve scrutiny
Motilal Oswal, in its May 4 report, drew a distinction the brokerage said the market consistently conflates when evaluating large system integrators’ pricing model announcements.
When large IT firms say they are moving to outcome-based pricing, they almost universally mean scope and price are defined around a business output and delivery milestones, not headcount. Outcome-contingent or gain-share pricing, where the fee is directly linked to actual performance improvement achieved, is a materially different and far harder commercial model.
What Motilal Oswal is watching over the next 6 to 9 months
The brokerage said it will track two signals in upcoming quarters: pricing commentary in managed services deals for any explicit mention of outcome-linked structures in large-cap results, and AI-native partnership announcements involving Indian technology vendors alongside evidence of short-cycle AI-led deal wins at meaningful scale.
Motilal Oswal’s analysis of application programming interface calls for Claude and token usage for OpenAI found that software engineering accounts for 50% of all API calls, and that 90% of the top 20 token users for OpenAI are new-age technology companies rather than traditional enterprises. This indicates AI is currently easier to deploy in greenfield environments, which are predominantly served by cloud-native and AI-native firms rather than the large Indian system integrators whose strength lies in complex brownfield enterprise deployments.
Conclusion
The brokerage concluded that near-term value capture remains skewed toward AI-native enterprises, with IT services vendors absorbing pricing pressure and growth visibility staying constrained well into FY27.
Disclaimer: This report contains specific stock ratings, price targets, and sectoral investment analysis that may impact financial decisions. The views expressed are based on market reports and brokerage analysis and do not constitute a direct offer or solicitation to buy or sell securities. Readers are advised to consult with a SEBI-registered investment advisor or a qualified financial professional before making any investment commitments based on these projections or valuations.
This disclaimer has been generated using AI to support user well-being and responsible content consumption.