No AI-led job apocalypse yet, but India must still stay cautious, says Economic Survey

Greater visibility into AI adoption has brought greater clarity on the nature of the technology itself, the Survey stated

• AI has not caused major job losses yet, says Economic Survey 2025-26
• Labour markets show slight differences between AI-exposed jobs and others
• AI adoption is rising, but usage is concentrated in high-income countries.

Did our AI summary help?

The 'doomsday scenario' for artificial intelligence (AI)'s impact on job losses hasn't arrived—at least, not yet.

The Economic Survey 2025-26, which was tabled in Parliament on January 29, states that early evidence has begun to temper some of the "extreme predictions surrounding AI’s near-term labour impact".

For instance, an Yale’s Budget Lab study indicates that the broader labour market in the United States has not experienced a discernible disruption due to AI while a study by Brynjolfsson, Chandar, and Chen (2025) highlights that the difference in job prospects between occupations highly exposed to AI and those with relatively low exposure is minor.

This provides some reassurance in the near term, especially for labour-abundant economies such as India, the Survey notes. However, it warns that this is no time for complacency, especially from a policymaker’s perspective.

"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 stated.

The Survey further 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'.

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 labor displacement—particularly in the middle- and lower-quartiles of the wage distribution.

The latest Survey reflects on 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"

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 percent of all usage in April 2025, adoption in upper- and lower-middle income countries has expanded to 22.5 percent and 18.7 percent respectively, it noted.

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