A national AI-training target can conceal a practical question: what work will people use that training to do? A course for an office team and one for a self-employed worker may share basic lessons, but their applications need not be the same.
An October white paper from Great Lakes’ Centre for Finance and Economics Research classifies 48.3% of covered urban employment as AI-exposed, against 15.0% in rural areas. “Exposed” means above-average overlap between occupational tasks and AI capabilities in the reference index. The complementary subset—work potentially assisted by AI—represents 15.8% of covered urban employment and 4.3% of rural employment.
The Breakdown
The analysis uses survey-weighted 2025 Periodic Labour Force Survey data for employed people aged 15–64, including self-employed and unpaid family workers. Street and related sales/service workers are excluded because a required task measure is missing.
Fixed measures derived from US occupational tasks are assigned to broad occupational groups. They cannot capture differences within occupations or actual Indian workplaces. The paper’s alternative-index comparison does not recalculate the urban/rural shares.
Why This Matters
For someone designing training, the useful next step is to examine the tasks behind the target audience. Which tools could help with those tasks? What equipment, language support and supervision would learners need? How would the programme establish that people can apply what they learn?
Those are planning questions, not interventions validated by this study. A lower measured exposure should not become a reason to exclude a community from training. Equally, a higher score is not enough to choose a course or promise an earnings increase.
The Big Picture
The International Labour Organization’s May 2025 update uses a different framework: task-level evidence, expert input and AI predictions are combined into four exposure gradients. It expects transformation to be more common than redundancy because most occupations still require human input. That global analysis does not independently confirm the Indian urban/rural estimates.
The Crunch
The practical test is whether a training plan connects a defined group of learners to useful work. Enrolment can show how many people joined. Demonstrated application is a separate question.