The Economic Survey has cautioned that India’s entry into the artificial intelligence era comes with “looming uncertainties” and structural asymmetries that could reshape labour markets and pose fundamental questions for the future of the Indian IT sector, even as the country seeks to harness AI-driven productivity gains.
The Survey flagged the growing concentration of market power in the hands of a few global frontier model builders, warning that control over critical AI inputs such as data, compute infrastructure and foundational models is increasingly restricted to a small set of firms, PTI reported. This, it said, raises concerns around technological dependence, market dominance and the resilience of global supply chains.
While India enters the AI phase with notable strengths, including a large talent pool and a vibrant digital ecosystem, the Survey noted that access to cutting-edge compute infrastructure remains limited and financial resources for training large-scale foundational models are scarce compared with global leaders. As a result, pursuing frontier model development as the core of India’s AI strategy would be “challenging”.
Instead, the Survey advocated a bottom-up approach to AI adoption that aligns more closely with India’s economic realities. It argued that value creation in AI need not be concentrated in a handful of large models or firms, and made a case for application-specific, smaller models tailored to defined use cases and sectoral needs.
On labour market implications, the Survey said that while emerging evidence offers some near-term reassurance for economies such as India, there is “no room for complacency”, particularly for policymakers.
“All in all, caution is still warranted as India attempts to solve the puzzle of AI and labour. This represents one of the most considerable looming uncertainties about the technology,” it said.
The uncertainty, the Survey added, extends to the evolving structure of the global AI ecosystem. Firms that once relied on India’s comparative advantage to handle large volumes of IT and services work may increasingly automate these tasks, raising questions about the long-term demand for traditional outsourcing models.
“It also raises a substantial question about the future of India’s IT sector,” the Survey said, warning that failure to adapt could risk “hollowing out” India’s core value proposition. Sustaining competitiveness, it said, would require a comprehensive evolution that fully leverages AI development and deployment.
The Survey also highlighted the global divide between frontier model development and application-led AI. The capability to design and train large foundational models remains highly concentrated among a few firms, which exert strong control over markets and place heavy demand on scarce resources, creating high barriers to entry.
Export restrictions on advanced processors needed to scale frontier models further compound the challenge for countries like India. This, the Survey said, creates a fundamental asymmetry in which most countries participate in AI largely as users, while a few shape the technology’s trajectory, standards and pricing.
Attempting to close this gap through heavy public spending would involve prohibitive fiscal costs and could prove unsustainable. The policy trade-off, therefore, lies between chasing frontier-scale models and deploying scarce resources towards domain-specific AI systems aligned with domestic priorities.
The Survey underlined that India’s challenge is not whether to adopt AI, but how to pace its diffusion so that labour augmentation is prioritised over displacement. Rapid, uncalibrated deployment could boost output but displace workers faster than the economy can absorb them, while delaying adoption to protect jobs risks locking firms into a low-productivity equilibrium.
Regulation is another area of uncertainty, with countries diverging in their institutional approaches to AI governance. For India, the Survey said, the task is to govern AI in a manner sensitive to domestic economic realities, ensuring that the benefits of AI diffusion are broadly shared across sectors and people.
It also stressed the need to balance openness with stewardship, so that economic value generated from domestic data and intellectual property accrues within India rather than being captured abroad.
Start a Conversation
Post comment