Coming of age can be traumatic. It involves a journey of discovery when youngsters come face to face with the adult world and its many challenges while coping with the inadequacies of their youth. It is when the threats of the real world seem so overwhelming that the time of growing up turns into a painful period. For the technology sector, which has been in a permanent state of adolescence, 2025 was the year of reckoning when the menacing new world of artificial intelligence (AI) and advanced avionics, for instance, provided a harsh learning experience for both innovators and the government. More so for the latter, which has wasted precious time in promoting the glories of a mythical past when India was supposed to have cracked all the sciences and pioneered everything from medical transplants to cosmic warfare. The past year showed us our place in the world on many fronts.
The most telling moment was the stunning debut of China’s DeepSeek AI chatbot, a large language model (LLM) whose prowess rivalled that of the top tech leaders like AI Open and Nvidia. India felt the impact more sharply, shaking the Ministry of Electronics and Information Technology (MEITY) out of its complacency like nothing else had. Within a fortnight, a galvanised government had invited proposals for the country’s own foundation models—these are large models that serve as a base for specialised AI applications that can be adapted to perform a range of tasks—and also asked private companies to reserve graphic processing units or GPUs with sufficient computing capacity for government led AI research. India suffers from an acute scarcity of GPUs.
Both invitations elicited a reasonably good response. MEITY garnered access to nearly 19,000 GPUs at subsidised rates, earmarked for foundational AI projects. This in turn brought in a rush of proposals from private entities for building their own models. It is still just a trickle. China is the global leader boasting a host of startups, thanks to a sustained focus on scientific and engineering breakthroughs that have come about as a result of an unbeatable combination: high-quality technical education and research institutes, and the meshing of academia with industry. So it is no surprise that China's startups have ventured into deep tech sectors like AI, semiconductors, robotics and 3D manufacturing, while India takes pride in startups that do not push the boundaries of technology.
Unexpectedly, it was a senior minister in the Narendra Modi government who drove home this uncomfortable point last year. At the grandiosely titled "StartUp Mahakumbh", Union commerce minister Piyush Goyal startled the convention by showing a slide that listed the difference in India’s and China’s approach to startup ventures. While India had food and grocery delivery services and applications for betting and fantasy sports, the Chinese were intent on electric vehicles and battery technology; semiconductors and AI; global logistics and trade; robotics and automation; renewable energy and space high-speed rail. In all sectors, the scale was so huge that the Chinese companies dominated the global market. Goyal’s comments were scathing. India’s startups, he said, were just businesses that are turning unemployed youth into cheap labour so that the rich could get their meals at home.
China, on the other hand has much to show. Apart from DeepSeek, it has MiniMax and other "AI tigers", which are among the top five models globally.
Unfair or not, Goyal’s take on the state of India’s tech startups provided a much-needed jolt that also identified China as a benchmark. China’s sustained technological advancements should have prompted the government to do something much earlier. Around the time India was plunged into chaos over a disastrous demonetisation exercise, China was preparing a blueprint for charting the country’s approach to developing AI technology and applications. A seminal document, "New Generation Artificial Intelligence Development Plan", setting out phased goals to make China a world leader in AI by 2030, was unveiled in early 2017. If the Bharatiya Janata Party (BJP) government had set about understanding this policy document instead of looking back—its fixation is the alleged misrule of Nehru and the Congress in the past 70 years—India would not be lagging so much in technology. It was only in 2024 that the US $1.25 billion India AI Mission was launched to build India’s core AI infrastructure. Managed by MEITY, it aims to make advanced tools more widely accessible by supporting startups such as Sarvam, which are developing foundation models in Indian languages and will, hopefully, apply AI to key sectors such as agriculture, education and healthcare.
AI is finally in the air these days. The fizz is due to the India-AI Impact Summit 2026 which opens in New Delhi on February 15. India has invited China to participate in the summit and it is seen as a sign that we are finally ready to learn from the leader. And there is much to learn. China is systematically integrating AI into various sectors such as healthcare, manufacturing, transportation and infrastructure and even governance through its AI+ Initiative. India, on the other hand, is just "hoping" to have a few foundational model applicants ready to be showcased at the summit, according to the official in charge.
But that has not stopped the government from its customary bombast. Catchphrases and lofty aphorisms from the past mark the goals of the AI Impact Summit. Themes for global cooperation, says the summit website, are based on the three sutras (ancient rules) of people, planet and progress—as pointless as it gets—which will act as the guiding principles for the seven chakras (energy points in the body) to enable concrete action. Confused? Never mind. None of the leading AI nations are likely to be impressed by this.
China’s extraordinary success came through long-term strategy, deep investment in talent and infrastructure. India, meanwhile, continues to encourage the exodus of its best talent to build US tech supremacy and is constrained by a severe shortage of computing power. There is a lesson here for Indian startups. When the US imposed strict export controls on advanced chips, China’s AI firms overcame the constraints through architectural innovation, efficiency and open-source collaboration. It resulted in extraordinary resilience and self-reliance, concepts that we are fond of plugging without taking concrete steps in this direction.
An analyst at the AI Insight Think Tank has noted that it is the core deficiencies in India’s programme that is prompting it to seek global cooperation.
This article was originally published in the the February 1-15, 2026 print edition of Down To Earth