I went into the discussion on India’s AI priorities with a familiar assumption. Technology has disrupted industries before. Economies have gone through recessions, businesses have adapted, new jobs have emerged and the workforce has gradually adjusted. Perhaps AI would find its own course correction too.
I came away less certain.
The essays and panel discussions at the launch of Strategic Priorities for India: Essays on AI Policy, an initiative by Digital Futures Lab and Samagata Foundation made me think harder about the scale of the change ahead. AI is not simply another technology for businesses to adopt, or an industry cycle that will eventually settle. Its implications reach India’s technological independence, the future of work, the choices open to young people and the country’s position in the global economy.
These questions cannot be left to industry professionals and think tanks alone. They need an equally considered response from government, one that keeps pace with the speed of technological change.
Two discussions stood out. One examined India’s technological autonomy in the age of AI. The other explored the impact of AI on India’s IT and ITES workforce and the policy responses that may be needed. Together they raised a larger question: is India preparing only to take part in the AI race, or also to manage its consequences?
Venkatesh Hariharan’s essay “India's Strategic Autonomy in the Age of AI” was an eye-opener. It brought into focus a risk that can be overlooked when conversations centre on innovation, productivity and growth: dependence on technologies we do not control.
Technology denial, vendor dependency and surveillance are not abstract geopolitical concerns. For a country like India, they raise basic questions about access, control and the ability to make independent decisions in a world that increasingly relies on AI.
One point that particularly caught my attention was the distinction between open source AI models and models that merely describe themselves as open source. The Open Source AI Definition, developed by the Open Source Initiative (OSI), establishes specific requirements for AI systems to qualify as genuinely open source. Only AI systems that meet this definition should be described as open source AI. This distinction is important because it protects the fundamental freedoms to use, study, modify and share these systems. Making model weights publicly available, for instance, does not automatically make an AI model open source.
The distinction matters even more when viewed through the lens of national strategic autonomy. If India is to build its AI capabilities on open technologies, it needs to understand what is genuinely open and what dependencies may remain beneath the surface.
The essay also makes the case for India to move beyond being a consumer of open-source technologies and become an active contributor to the global AI ecosystem. Independence need not mean building everything in isolation. It can also mean taking part in shared ecosystems, contributing to their development and having the capability to adapt them to our needs.
Parminder Jeet Singh’s essay “AI as Digital Public Infrastructure” offers another direction. It raises questions about how AI capabilities might be developed and made accessible in the public interest. Could shared infrastructure help Indian businesses, researchers, public institutions and smaller enterprises use AI without depending entirely on a handful of private providers? What would India need to build, govern and maintain to make that possible?
These questions deserve serious attention. India has already taken steps to build its AI ecosystem. The larger task is to ensure that these initiatives add up to a coherent national approach that addresses technological capability, autonomy and public interest together.
Amrita Sengupta’s discussion, “AI, Labour, and the IT/ITES Sector,” brought the employment question into sharper focus.
India’s IT services industry has long been a major source of jobs, exports and economic opportunity. AI could change the economics of this industry by automating tasks, raising productivity and altering the skills clients expect from service providers.
The precise impact on employment remains uncertain. Productivity gains do not automatically translate into job losses, just as the creation of new roles does not guarantee that those displaced will be able to move into them. The essay examines several possible labour market scenarios and argues for better evidence, reskilling, career planning and coordinated policy responses.
What stayed with me was the need to prepare before the full impact becomes apparent. If we wait until existing roles disappear or skills become obsolete, workers may have too little time to adapt.
My concern is not only whether AI will replace certain jobs. It is whether we are identifying the jobs that may emerge, understanding the skills they will need and preparing people to move into them.
AI is increasingly a horizontal technology, affecting work well beyond the IT sector. Its applications will extend across manufacturing, healthcare, financial services, logistics, education and other industries. India’s response to the employment challenge therefore needs to look beyond training more AI engineers. It must also consider how people in different occupations can work alongside AI, how new technical roles will develop and where other employment opportunities may arise.
This calls for closer coordination between industry, educational institutions, skilling bodies and government, and for a better understanding of the jobs India will need, instead of simply training people in the skills currently in demand.
During the Q&A session, a participant suggested that India place greater emphasis on vocational training in fields such as electrical work, through Industrial Training Institutes (ITIs). The suggestion made sense for an economy that will keep needing skilled technical professionals even as AI changes the nature of many white-collar jobs.
Aastha Kapoor raised a counterpoint about how far the curricula in these institutions lag behind current requirements, and how urgently they need upgrading.
This brought another dimension of India’s skilling challenge into focus. Improving access to vocational education is one thing. Ensuring that the education is relevant, the qualification is valued and the career is considered desirable is quite another.
One example shared during the session involved the son of a domestic worker who had completed Class 12. He might choose a BA degree over an ITI qualification, even though vocational training could offer better earning prospects. A conventional degree may carry greater social acceptance, while skilled technical work is often seen as a lesser option.
I found the example telling because it exposes a contradiction in how we think about education and employment. We speak of a shortage of skilled workers, yet the pathways that could prepare people for these roles do not enjoy the same standing as conventional degrees.
Two problems need attention here, and they have to be solved together. The first is the quality and relevance of training. Vocational institutions need current curricula, practical learning, industry participation and credible pathways to employment. The second is the value society attaches to the qualifications they offer.
Updating an ITI syllabus will not be enough if the qualification continues to carry lower status. Changing perceptions without better training and real employment prospects behind them will not solve the problem either.
India needs to make skilled technical work a credible and aspirational career choice, supported by qualifications that employers recognise and opportunities for continued learning and progression. This is an education question, and it is also a question of how India prepares its workforce for an economy in which technological change will touch jobs at every skill level.
Having spent more than twenty years in enterprise technology marketing, I see three practical implications for companies that build or sell AI.
First, claims about openness will be examined more closely. Buyers and regulators are starting to ask what “open” actually covers, including model weights, training data and the freedom to modify. A company that uses the word loosely will find its claims harder to defend later. The description should match what the product genuinely allows.
Second, the workforce story belongs in how a company presents itself to the market. IT services firms and global capability centres will be asked how AI affects their people, and clients, employees and policymakers will all listen closely to the answer. Companies that speak plainly about reskilling and career pathways will earn more trust than those that stay silent until questions arrive. Third, perception shapes the supply of talent. If skilled technical work carries little social standing, hiring pipelines will suffer however good the training becomes. Employers have a part in changing how these careers are seen.
What connected these discussions for me was the urgency of the response required. India’s AI ambitions cannot be measured only by investment, infrastructure or the number of models it builds. We also need to ask whether institutions, industries and people are being prepared for what this technology will bring.
One area worth examining is the systematic identification of emerging AI-related roles across industries. Understanding which tasks are changing, which new capabilities are needed and where employment opportunities may emerge could help India align training and reskilling efforts with future demand. This requires better labour market data, active industry participation and training programmes that can keep pace with changing requirements. It also requires coordination across government departments, because the implications of AI reach beyond technology policy into education, labour, industry and economic development.
The task is to bring these efforts together early. If we wait until job losses are evident or skills have become obsolete, the cost of catching up will be higher for individuals and for the economy.
I began with the belief that, like earlier economic disruptions, AI would eventually find its own course correction. I still think markets and societies will adapt. What I now doubt is that adaptation will arrive on its own and in time, particularly when the choices made today will shape the opportunities open to an entire generation.
Markets will do much of the work, and companies will carry a large share of it. The pace and spread of this change, though, call for preparation that no single party can manage alone. Government can set direction and support shared infrastructure. Industry can say plainly how jobs and skills are changing. Educators and researchers can supply the training and the evidence. Each does better when the others are moving the same way.
The technology will continue to advance. The open question is whether our institutions, policies and education systems are ready to keep pace with it, and that readiness is something we can begin building now.
The author is Founder and Director at Whiterays, a technology communications and marketing advisory firm. Reach out to the Whiterays team at contact@whiterays.com