Indian organizations are experiencing a significant disconnect between AI adoption and measurable returns, according to a new ISACA report based on polling over 3,400 digital trust professionals, including 265 from India. While 86% of employees in India use AI at work, only 35% say the technology has delivered returns that met or exceeded expectations, representing a higher adoption rate than the global average of 22%. This disconnect points to a deeper problem: organizations are deploying AI faster than they can govern it or understand its impact. Keith Bloomfield-DeWeese, Senior Manager of AI Product Development at ISACA, explained that "Return on investment in AI doesn't arrive on schedule; it's the result of sustained investment in people, processes, and governance structures."
Indian IT services companies are rapidly adopting innovative pricing models for artificial intelligence work as soaring API and usage-based costs are forcing companies to rethink spending. According to The Economic Times, Latentforce's token cost has skyrocketed from ₹20,000 per month six months ago to ₹2-3 lakh per month currently, while Uber's chief technology officer recently stated in an internal memo that the company burnt its 2026 AI budget in four months. This cost escalation is prompting enterprises to hire humans instead of deploying AI, with companies like Latentforce's clients in the knowledge services sector hiring people rather than deploying AI as the former is cheaper currently. As reported by The Information, there have been multiple instances of employees exhausting a year's worth of AI budget within months, creating a paradox where demand for AI productivity tools is higher than ever, but AI costs are driving higher than human capital costs.
Indian IT services companies are rapidly adopting innovative pricing models for artificial intelligence work, moving away from traditional bill-by-the-hour models. According to reports from Mint, EPAM Systems is factoring AI token costs into deal structures, while Tech Mahindra, India's fifth largest IT firm, is exploring pricing linked to token consumption. Cognizant Technology Solutions Corp has introduced a rate card for AI usage, and Coforge Ltd has launched a fixed monthly subscription for AI use, offering clients a buffet of over 130 pre-built AI agents. The shift comes as large AI enterprises such as Anthropic and GitHub move from current flat subscription structures to usage-based pricing, with GitHub transitioning to usage-based billing starting June 1 as absorbing escalating computing costs becomes unsustainable.
Indian employees are primarily using AI for productivity enhancement and automation, according to the ISACA report. Indian employees most often use AI to increase productivity (56%), automate repetitive tasks (55%), create written content (51%), and analyze large amounts of data (42%). Eighty-two percent of respondents said AI skills are very important to their profession, yet thirty-five percent said their organizations now train all employees on AI, up from 22% in 2025. The report reveals significant gaps in AI governance, with forty-nine percent of Indian organizations now having a formal, comprehensive AI policy, up from 32% in 2025 - above the 38% global rate. However, twenty-three percent of organizations have only limited policies, and 20% have no active policy at all, creating challenges for consistent measurement and resource allocation.
Organizations are planning significant expansion of AI-related roles despite current governance challenges. Fifty-seven percent of Indian respondents said their organization will increase AI-related jobs in the next 12 months, up from 46% in 2025. This expansion signals where the industry is heading, with organizations committing resources to AI development. However, the ISACA report shows that when asked about returns, Indian respondents gave scattered answers - twenty-one percent said it's too early to tell, another 21% cited limited ROI so far, and eighteen percent were unaware of any returns at all. Companies for which AI tools are a mainstay are optimizing internal AI usage and tweaking revenue models, with Latentforce exploring optimizing AI model usage by a mixture of open source and frontier models and implementing agentic guardrails to ensure tokens are used for approved projects.
