The IndiaAI Mission is building technological capabilities through investments in computing infrastructure, indigenous foundation models, datasets, research, startups and Centres of Excellence. As these capabilities develop, a second task is becoming urgent: preparing India’s workforce for an economy in which AI is changing how work is done. Proposals for an India AI Talent Mission and a National Workforce Transition Framework recognise the need for greater coordination across skills, education, industry and labour markets.
The question is what an operational workforce-transition system should look like, where it should sit, and how quickly it can begin generating results.
The Transition Is Already Underway
AI is altering the tasks within jobs; occupations may survive even as the work within them changes. This matters because education and skilling systems train largely for occupations and qualifications, while AI is changing the tasks they contain.
The World Economic Forum estimates that 39% of workers’ existing skill sets globally will be transformed or become outdated between 2025 and 2030. Training designed today can therefore become misaligned with employers’ needs relatively quickly. Businesses cannot wait for education systems to catch up, yet universities cannot redesign curricula every time a new AI application appears.
The institutional problem is one of timing and translation: detecting changes in work early enough and converting those signals into decisions about curricula, training, hiring and job design.
When Firms Adapt at Different Speeds
Large technology companies and global capability centres can build internal systems to identify emerging skills, retrain employees and reorganise work. Many smaller firms cannot justify the same investment. India could therefore develop pockets of world-class AI adoption without wider diffusion, with productivity gains concentrated among firms able to redesign work.
A workforce-transition system can reduce this gap by making intelligence on changing tasks, skills and job roles available beyond firms capable of generating it themselves. Workforce transition is therefore a productivity issue as much as a labour-market issue.
From Changing Work to Action
India needs a mechanism that continuously detects which tasks are changing and translates those changes into skill and training responses. At present, the relevant functions are spread across technology, labour, education and skilling institutions, with no clear mechanism connecting them.
A dedicated coordination unit within the proposed India AI Talent Mission could close this gap without replacing existing institutional responsibilities. It could track changes in tasks and skills, convene employers, universities and training providers, initiate pilots and feed evidence to the institutions responsible for curricula, training and labour-market policy.
This would also connect workforce policy more directly to AI adoption. The objective would be to identify changes in work while they are occurring, rather than redesign training after skill gaps have become entrenched.
Map, Test, Then Scale
The system should begin in a small number of sectors where AI-driven changes in work are already visible. The first task is to map which tasks are changing, what new skills they require and where existing training is becoming outdated.
Employers, universities and training providers can then develop transition pathways for priority occupations and test them with firms and workers. These could include revised curricula, short courses, apprenticeships or mid-career training, with States adapting the pathways to their industrial and employment structures.
The results should determine what scales. Pilots can establish whether particular interventions improve worker mobility, retention or productivity, allowing successful approaches to expand while ineffective ones are modified or dropped.
India should build its workforce-transition system alongside its AI infrastructure, rather than wait for labour-market disruption to make adaptation unavoidable.