In 1811, armed with hammers and desperation, a band of English textile workers stormed factories to smash the mechanised looms that threatened their livelihoods. These ‘Luddites’ weren’t resisting technology itself—they were fighting an economic shift that had no place for them. Fast forward two centuries: the anxiety remains, though the machines have changed.
In 2016, when AlphaGo dismantled world champion Lee Sedol at the ancient game of Go—long thought to be beyond machine capabilities—tech writer Kevin Kelly issued a chilling prediction: “Before AI, a job was safe if it required intelligence. Now, those are the first to go.” The Economic Survey 2024-25 wrestles with this same conundrum: will artificial intelligence be a force for empowerment or an architect of obsolescence?
The survey dedicates a chapter to the impact of AI on employment, portraying it as both a challenge and an opportunity. However, the analysis is rooted in theoretical optimism, assuming that institutional mechanisms will adequately mitigate disruptions. While acknowledging risks such as job displacement, inequality and the concentration of AI benefits among a few firms, the survey provides solutions that may downplay the account for the scale and speed of AI-driven changes.
Today, AI is not only taking over repetitive jobs, but also jobs that require critical thinking. In healthcare, AI systems like DeepMind can diagnose diseases from medical images more accurately than radiologists, while Watson provides personalised oncological treatment recommendations by analysing vast amounts of clinical data. In criminal justice, tools like COMPAS assess recidivism risk more consistently than judges, and predictive policing algorithms help law enforcement allocate resources based on crime patterns. In education, AI-driven platforms like DreamBox and Khan Academy tailor learning experiences to individual students. In financial services, AI models power algorithmic trading for faster, more accurate market predictions, and fraud detection systems block suspicious transactions in real-time.
The penetration of AI into India’s job market is expected to induce significant structural shifts, particularly in labour-intensive service sectors. An IIM Ahmedabad survey reveals that 68 percent of white-collar professionals anticipate their roles being partially or fully automated within five years, while 40 percent foresee skill obsolescence. McKinsey Global Institute estimates that up to 30 percent of work hours in Europe and the US could be automated by 2030, suggesting similar trends in India’s IT and process outsourcing industries, which employ 5.4 million people.
Job losses in IT services and customer support could exacerbate unemployment, given that services contribute half of India’s GDP and remain the largest employer outside agriculture. But adoption is already accelerating, as evidenced by PhonePe’s 60 percent reduction in customer support staff and ICICI Bank’s AI-led automation of credit underwriting and risk modelling.
The survey provides a forward-looking framework for AI integration, emphasising skilling, education and institutional reforms to mitigate labour displacement and enhance productivity. It rightly argues that AI should be a labour-augmenting force rather than a substitute, particularly in high-demand service sectors.
By shifting the focus from narrow AI-specific training to foundational skills like critical thinking, adaptability and creativity, the survey recognises the long-term importance of a dynamic and resilient workforce. It also highlights the need for demand-driven skilling ecosystems, ensuring that automation-induced job losses are counterbalanced by employment creation in emerging roles.
The assumption that India has “time to build institutions” before AI-driven disruptions become widespread is problematic. AI’s rapid adoption, particularly in services-heavy economies like India, will likely outstrip the pace at which enabling, insuring and stewarding institutions can be established and made effective. Institutional reforms—especially those aimed at skilling, reskilling, and establishing safety nets—historically take years, if not decades, to show measurable impact, whereas AI-driven automation is advancing at an exponential rate, driven by market forces, global competition and cost-cutting imperatives. Unlike previous industrial revolutions, where technology complemented human labour before displacing it, AI directly substitutes white-collar cognitive and decision-making tasks, meaning that job losses will not be confined to low-skill workers alone.
India’s existing institutional shortcomings exacerbate this challenge. The Kaushal Vikas Yojana, the country’s primary skilling programme, has not been as successful as it should have been due to weak industry linkages, poor quality training and a lack of alignment with market demands. Expecting a rapid institutional pivot towards AI-specific upskilling is unrealistic, given limited fiscal space, administrative inefficiencies and the decentralised nature of education.
The sectoral composition of the labour market accelerates the risk. The IT and outsourcing sector is particularly vulnerable to large-scale automation as AI models begin to handle customer service, data processing and even software development. Unlike manufacturing, which required massive capital investments and physical infrastructure, AI adoption is software-based and can be implemented rapidly. The outsourcing sector in the Philippines is already seeing significant AI-driven job losses, and India will not be immune.
Unlike past transitions, AI adoption will be sudden and widespread, leaving little room for reactive policymaking. So it requires urgent, proactive action now.
(Views are personal)
Aditya Sinha
Public policy professional
(On X @adityasinha004)