Should AI be taxed?
What do you think?
Mohammad Danish
7/13/20263 min read
Senator Elizabeth Warren’s proposal to tax AI companies, especially those running large data centers, should not be copied blindly by India, but it should definitely make India think. AI is not just another software wave. It will consume land, electricity, water, chips, talent, public infrastructure and human-created data at a massive scale. If India only becomes a market for foreign AI tools and a low-cost data-center destination, it will miss the larger opportunity. But if India designs the right policy now, AI growth can become a national wealth engine that funds jobs, research, skilling, clean energy, semiconductor capability and new technology exports.
India should not introduce a punishment-style AI tax that discourages investment. The smarter approach would be an “AI Infrastructure Contribution Framework.” Large AI data centers, hyperscalers and high-compute AI companies could contribute through a graded levy linked to compute scale, electricity intensity, water use, carbon footprint and revenue generated from Indian users. Smaller startups, research labs, universities and Indian MSMEs should be exempt or subsidised. The principle should be simple: do not tax innovation; tax excessive extraction from public resources.
The money collected should not disappear into the general budget. India should create a ring-fenced “National AI Future Fund.” This fund should finance five things: AI skilling, job creation, Indian-language AI models, export-ready AI startups, and green infrastructure for data centers. The goal should not merely be revenue collection. The goal should be to convert AI’s private gains into public capability.
India must also address the growing perception that it has “missed the AI bus.” This perception is emotionally powerful but strategically incomplete. India was never the country that typically invented the foundational layers of global technology. It did not write the world’s dominant programming languages, but it produced some of the world’s finest programmers. It did not invent every major enterprise software platform, but it became the world’s most trusted technology execution partner. India’s strength has rarely been in creating the first abstract breakthrough. Its strength has been in solving known, complex, scaled problems better, cheaper and faster than most countries.
That is exactly where India can win in AI. The world will not only need foundational AI models. It will need AI deployment, AI governance, AI cybersecurity, AI integration, AI compliance, AI-enabled manufacturing, AI-led customer service, AI agents for enterprises, AI tools for healthcare, education, logistics, agriculture, banking and public administration. These are not theoretical problems. These are real operating problems. India understands operating complexity at scale.
To mitigate the risk of being seen as a passive AI consumer, India should take visible and practical steps. First, it should build national AI compute capacity accessible to startups, universities and smaller companies. Second, it should create sector-specific AI missions for healthcare, agriculture, legal services, manufacturing, logistics and education. Third, it should push Indian IT services companies to move from manpower-led delivery to IP-led AI platforms. Fourth, it should incentivise AI hardware, chip design, server assembly, cooling systems and data-center equipment manufacturing. Fifth, it should make AI skilling as aggressive as the IT training boom of the 1990s and 2000s. Sixth, it should create procurement policies where government departments buy Indian AI solutions wherever possible, giving local companies a home market before they compete globally.
India should also negotiate better with global AI companies. Any large AI data-center approval should include commitments on renewable energy, water recycling, local employment, Indian-language AI development, local procurement, cybersecurity and startup partnerships. Companies that contribute to India’s AI ecosystem should receive incentives. Companies that only consume Indian resources and extract profit should contribute more.
The larger question is not whether India should tax AI. The question is whether India will treat AI as a passive consumption wave or as a sovereign economic opportunity. India does not need to feel ashamed that it did not create the first wave of foundational AI models. Its opportunity lies in solving the problems AI will create for the world — software problems, manufacturing problems, compliance problems, workforce problems, infrastructure problems and deployment problems. Just as India became the global supplier of software talent without inventing programming itself, it can become the global supplier of AI-era solutions without owning every foundational model.
If done well, India will not miss the AI bus. It may simply board it from a different door — not as the loudest inventor, but as the most dependable builder, integrator, supplier and problem-solver of the AI economy.
Journey well taken
Sharing life insights on marketing and motorcycling adventures.
Connect
© 2026. All rights reserved.