Economic Survey: No AI-led Job Apocalypse Yet; Strategy Must Prioritise Capacity, Coordination

The Survey, tabled by Union Finance Minister Nirmala Sitharaman, said that the expanded use of AI has led to uncertainities. “Artificial Intelligence does not confront India with a single policy question, but a series of choices that must be made under conditions of heightened uncertainty and resource constraints,” it said. 

The survey noted that while labour may be complemented in the near term as organisations work to incorporate AI into their tasks, productivity gains from this 'augmentation' phase have a 'ceiling'.

The Survey’s comments on India’s AI roadmap come ahead of the country hosting the AI Impact Summit 2026 next month. The Summit will be held in New Delhi from February 15 to 20, with the forum coming to the Global South for the first time.

This marks a subtle shift in tone from last year’s Economic Survey, which had warned that AI was poised to surpass human performance in critical decision-making, potentially triggering widespread labour displacement, particularly in the middle- and lower-quartiles of the wage distribution.

Possible AI Failure 

The Economic Survey, on the contrary, warned that AI failure may lead to broader financial contagion. It flagged risks to the global growth outlook, just like the dot-com bust, if the AI-led investment cycle fails to translate into anticipated productivity gains

"If the AI boom fails to deliver the anticipated productivity gains, it could trigger a correction in overly optimistic asset valuations, with the potential for broader financial contagion. Additionally, a protraction of trade conflicts would weigh on investment and further weaken the global growth outlook. These forces collectively suggest that downside risks to global growth remain prominent, although a fragile stability holds for now," it said.

Building Capacity

The Survey stated that India should build coordination first, capacity next, and binding policy leverage last, allowing institutions and markets to co-evolve.

The latest Survey reflects this shift, noting that when AI was last examined in early 2025, the conversation was still dominated by conjectural possibilities.

"While AI was already visible in some areas as a productivity-enhancing tool or embedded within some service platforms, its broader economic implications remained largely conjectural," it noted. "One year on, AI is no longer a distant or speculative technology. It is increasingly being adopted, even if in an experimental capacity, in organisations around the world," it added.

AI Adoption

Greater visibility into AI adoption has also brought greater clarity on the nature of the technology itself. The survey said, "Over the past year, it has become evident that while the use of AI tools can be widespread, the frontier of AI remains highly concentrated."

To be sure, the utilisation is still skewed towards high-income countries, which accounted for 58.4% of all usage in April 2025. Adoption in upper- and lower-middle income countries has expanded to 22.5% and 18.7% respectively, it noted.

Innovations and continuous improvement in AI capabilities are also driving firms and startups to develop ways to apply AI to solve real-world problems, the Survey said.

AI Strategy

It said that the first phase of India’s AI strategy should focus on operationalising already announced institutions and aligning incentives to enable experimentation.

“Policy should enable bottom-up innovation by expanding the reach of the existing shared infrastructure under the IndiaAI Mission. This includes a government-hosted community-curated code repository and pooled access to public datasets, facilitated by initiatives already underway to enable shared access to computing infrastructure. A clear focus on application- or sector-specific, small and open-weight models will enable efficient resource utilisation,” the Survey read.

It further recommended a clear focus on application- or sector-specific, small and open-weight models to enable efficient resource utilisation. It also suggested that AI regulation should be formalised on a risk-based and proportionate basis, with obligations for AI companies codified.