The India AI Impact Summit 2026 was unfortunately marred by allegations of plagiarism, leading to unseemly optics. The substance, in any case, was always a chimera, a delusion erected on foundations of sand. The chasm between our aspirational rhetoric and our investment reality looks unbridgeable.
The Union Budget 2026’s allocation of a mere Rs 1,000 crore for the IndiaAI Mission is, in the context of global capital flows into this domain, tokenism. It is a figure that speaks not of a national mission, but of a pilot project. It is not even enough to subsidise the compute costs for a significant portion of our research community, let alone build the kind of large-scale GPU clusters that would allow Indian researchers to train models comparable to GPT-6 or Claude 4.5.
By way of comparison, China spent USD 98 billion on AI in 2025 out of which the government spend was USD 56 billion. Five US-based private companies — Amazon, Google, Microsoft, Meta and Oracle — will collectively invest USD 700 billion in 2026. The Government of India will spend 1.2 billion over five years on its AI Mission.
The message from the finance mandarins is clear: Artificial intelligence, for all the ministerial speeches dedicated to it, remains just another sector, not a core strategic priority. The greatest risk we face is not an AI investment bubble, but insufficient investment, and consequently ceding the technological future to competitors like the US, China and Europe. Yet, our response to this existential challenge is a budget that has been effectively halved from previous projections, a move that signals to the world, and more importantly, to our own innovators, a comfortable acceptance of the slow lane.
This chronic underfunding has a corrosive effect that permeates the ecosystem. The IndiaAI Mission, with its stated goal of building sovereign capabilities, finds its ambition directly contradicted by the ground reality.
Nearly three out of four Indian AI deployments rely on Western proprietary models accessed via APIs. A staggering 65 per cent of GPU computing power among these start-ups is dedicated to inference (using models to complete tasks) while a paltry 21 per cent goes toward training new models. The argument that value lies in the application layer is a convenient rationalisation for our failure to compete at the foundational level.
The talent pool, too, is a casualty of this skewed investment. While MeitY mandarins underscore that global giants are looking to India to hire AI engineers, we must interrogate the nature of this demand. The shortage is not in machine learning theory, but in “AI Ops” — the practical, high-level expertise required to evaluate, optimise, deploy, and monitor complex AI systems in production.
Our education system, starved of research funding and world-class computational infrastructure, is producing a workforce that is highly skilled for service roles but ill-equipped for the frontier work of model development. We are training the global AI economy’s middle management, not its C-suite.
Fiscal timidity becomes even more alarming when viewed through the lens of global competition. The Stanford HAI Global AI Power Rankings consistently places India in the second tier, trailing not only the US and China but also a host of European and Asian economies.
The World Economic Forum’s recent ranking places us eighth in AI investment, a position that sounds respectable until one comprehends the gulf that separates the top three from the rest of the pack. The rhetoric of “AI for All” collides with the reality that “All” includes virtually no one when it comes to foundational investment.
The contrast with the global frontrunners is more than a matter of scale. The US and China are engaged in a high-stakes competition funded by hundreds of billions of dollars, a fact underscored by the patent race. While NASSCOM reports that over 86,000 AI patents were filed in India between 2010 and 2025, the Forbes analysis of the global AI patent race clarifies the distinction: China outpaces the US in sheer volume of patents, but that does not equate to winning the AI war.
Many patents are defensive, incremental, or filed for tax and compliance reasons. The real measure of sovereignty is the ability to build and train state-of-the-art models on homegrown infrastructure using sovereign data. On that front, the Carnegie Endowment’s recent analysis correctly identifies the missing pieces in India’s AI puzzle: Talent, data, and R&D funding.
The government has based its entire AI vision on India’s rapid adoption of digital payments. However, adoption of finished technology is not the same as ownership of the means of production. India’s UPI is a world-class digital public infrastructure, but it was built on top of underlying technologies and protocols that are globally available. AI is different. The models themselves are the product, and the countries that control the most advanced models will set the terms for everyone else. We become the perfect market for others’ innovations, our data and our market size fuelling their growth while we remain on the periphery of value creation.
So, what, then, was the purpose of a grand AI summit? Was it to provide a platform for our start-ups to network with the Western API providers they are already dependent on? Was it to showcase our “vibrant ecosystem” of consumers of technology, rather than producers? Was it a gathering where the real players — OpenAI, Anthropic, Google — politely acknowledged our market potential while their IP continues to form the bedrock of our digital economy?
The summit’s real value does not lie in the myriad photo opportunities or lofty declarations, but in its potential to catalyse a fundamental rethink of our AI fiscal strategy. The potential is real; the opportunity to build India into a global AI powerhouse exists. The adoption of AI agentic workflows among Indian developers shows the hunger and willingness to experiment. But we are asking our start-ups to run a marathon on a diet of crumbs. The global AI race is not a sprint; it requires sustained, massive investment in data centres, in research grants, and in PhD programmes.
If this is not prioritised, the summit will be remembered not as a turning point, but as a poignant misfortune of a nation that mistook a conference for a commitment. The world will smile for the camera, and quietly note that India, for all its grand ambitions, has chosen to remain a consumer in an age of creators.
The writer is a lawyer, third-term MP, and former Union Information and Broadcasting Minister
