DeepSeek has triggered a dramatic shift in the global AI race: it transformed the previously dominant framing of a China-U.S. AI rivalry into a broader global competition, especially among emerging powers in the Global South. This shift is particularly evident in India’s case.
DeepSeek’s overnight success significantly reshaped New Delhi’s perception of the AI landscape, resulting in concrete steps aimed at strengthening its position in the global AI competition. There have been notable shifts in India’s strategies and initiatives designed to foster the national AI innovation ecosystem, enabling it to compete with the United States and China through active engagement with Indian startups and entrepreneurs.
After DeepSeek: India’s New Perception of the AI Race
DeepSeek has fundamentally reshaped India’s perception of the AI race, sparking a wave of new debates. Chief among them is a renewed discussion about India’s relative position in global AI development.
Why is India lagging behind the U.S. and China in AI? Why hasn’t India produced AI advancements comparable to China’s DeepSeek or the United States’ ChatGPT? These concerns have become central to nearly all AI policy debates in India since the rise of DeepSeek. India’s IT Minister Ashwini Vaishnaw has frequently encountered these questions in recent public discussions. In other words, DeepSeek has ignited a broader critique of India’s positioning in the global AI landscape.
Concerns are growing that India risks becoming a digital colony of the U.S. and China, as both countries rapidly advance AI technologies that increasingly shape India’s technology landscape. This concern has been articulated by many policy experts, including Amitabh Kant, former CEO of India’s apex public policy think tank, NITI Aayog, and the G-20 Sherpa during India’s presidency. Writing for The Indian Express, Kant argued that DeepSeek symbolizes China’s advancements in AI, while ChatGPT reflects U.S. technological progress. Meanwhile, India remains primarily a user of these AI systems. This state of affairs raises critical concerns about India’s AI sovereignty and its capacity to develop homegrown AI technologies aligned with its cultural and economic priorities.
DeepSeek’s success also played a key role in shifting India’s AI strategy from a domestically focused approach to one with international and geoeconomic ambitions. India is now seeking to compete globally in AI development rather than concentrating solely on the implementation of AI tools within the Indian tech landscape.
Before DeepSeek, India saw AI primarily as a tool for domestic economic growth, particularly for achieving its targets of reaching 10 percent GDP growth by 2023 and creating a $1 trillion digital economy. However, these goals were more about leveraging existing tools domestically than about international competition or AI innovation. Now India is scaling up plans to compete at the global level, viewing AI as a critical industry from a geoeconomic, rather than simply economic, perspective.
More importantly, DeepSeek has instilled a sense of confidence among Indian policymakers that local startups and smaller investments can indeed drive AI innovation. Shortly after its release, Vaishnaw, India’s IT minister, hailed DeepSeek’s success, underscoring the fact that even smaller companies with relatively limited funding could compete with the heavily financed AI models produced by the global tech giants. His comments were seen as an indirect critique of Sam Altman, the chief executive of OpenAI, who visited India for the first time in 2023 and suggested that India should not attempt to build foundational AI models due to the high costs and resource-intensive nature of such technology
After the launch of DeepSeek, a new wave of enthusiasm emerged, buoyed by the belief that India can achieve AI breakthroughs with limited investments – similar to its success with the Digital Public Infrastructure model, which was not initially appreciated but was later adopted by other nations. This sentiment has been echoed by many policy advocates, who argue that DeepSeek has proven that Indian startups can compete with global AI tech companies like OpenAI. As they put it, “The AI arms race isn’t about money. It’s about hunger.”
The Indian State as a Financial Resource Provider
Against this backdrop, the Indian government has expanded its strategic role in strengthening the nation’s AI innovation ecosystem. By taking a more active leadership position, it aims to compete with other global AI leaders.
Following DeepSeek’s launch, governments around the world significantly increased their financial commitments, underscoring the strategic importance of AI investment. The Indian government’s latest budget, presented to Parliament shortly after the launch of DeepSeek, substantially increased the allocation for the India AI Mission – initially launched in 2024 – from approximately $66 million in the 2024-25 budget to around $240 million in the 2025-26 budget, an almost four-fold increase.
Moreover, as discussed below, the government announced various direct and indirect financial support measures aimed at bolstering the Indian private sector. Special provisions were also included in the Union Budget to develop key AI infrastructure and provide subsidies to startups for accelerated growth. These developments underscore the Indian state’s growing prominence as a significant financial supporter of AI development, especially in the post-DeepSeek era.
New State-Business Relations: Embarking on a New Type of Startup Capitalism
Beyond increasing the state budget for the AI mission, the Indian government has assumed a direct role in fostering AI development capable of competing with leading AI powers such as the United States and China.
Shortly after the launch of DeepSeek, the Indian government settled an ongoing debate within India’s policy and business communities by officially committing to the development of an indigenous AI foundation model. On January 30, 2025, the government invited proposals from Indian startups and enterprises to develop large multimodal models, large language models, and small language models specifically tailored to India’s needs.
As part of this initiative, the government has pledged essential support, including financial backing, access to high-quality datasets, and AI infrastructure such as subsidized computing power. This move marks a significant shift in India’s AI strategy, ushering in a new era of startup capitalism where the government plays an active role in AI development by equipping firms with the necessary resources to drive innovation and competitiveness.
India’s approach to startup capitalism presents a new model, differing from both the United States’ privately driven ecosystem and East Asia’s state-led initiatives. Unlike the privately-driven startup capitalism in the U.S., the Indian government has proactively supported startups by providing significant financial resources, data access, and infrastructure support. In this sense, India’s startup capitalism closely aligns with the East Asian approach, notably seen in countries such as China, South Korea, and Japan, where the state actively directs technology development.
Despite this similarity, a crucial difference remains. Whereas East Asian models typically encourage startups to collaborate with established large corporations to bolster national tech giants, India’s version of startup capitalism prioritizes independent growth, ensuring that startups retain autonomy while benefiting from state-backed support.
This innovative approach to startup capitalism is particularly evident in India’s strategy for achieving a homegrown foundational AI model. The government has committed to providing direct and substantial funding to compete with global platforms like DeepSeek and ChatGPT. The state plans to offer equity investments, positioning itself as a potential stakeholder in these startup ventures, but the central government has also allowed AI developers to seek co-financing from venture capitalists and regional government grants. In addition to providing funding, the government will supply computing resources at significantly subsidized rates and ensure data accessibility, as elaborated below, underscoring India’s unique strategic approach.
Securing Critical Infrastructure for India’s Position in the AI Race
GPU chips have emerged as the most critical AI infrastructure, providing the computational power required for rapid data processing and sophisticated neural networks. In response to surging demand and shortages, tech powers across the Indo-Pacific, including South Korea, Japan, and China, have prioritized securing their GPU supplies.
Following the DeepSeek revolution, India has also accelerated its efforts to strengthen the GPU supply chain to ensure that Indian startups remain competitive in AI development. In this context, India has committed to securing more than 18,000 GPUs, a significant increase from the 10,000 GPUs available in the pre-DeepSeek era. This commitment reflects India’s ambition to enhance its computational power and compete with the United States and China in AI advancement.
As noted by Vaishnaw, India has secured a number of GPUs far above the computational capacity of DeepSeek and comparable to that of ChatGPT: “DeepSeek AI was trained on 2,000 GPUs, ChatGPT was trained on 25,000 GPUs, and we now have 15,000 high-end GPUs available.” Furthermore, the government has pledged to make available a wide range of chips, including Nvidia A100 and H100 models, which are difficult for China to access due to export controls imposed by the United States. Indian authorities have selected 10 Indian companies to produce and procure GPUs, strengthening the country’s AI infrastructure for future advancements.
The objective is to provide computational power to Indian startups at a significantly lower cost – under 100 rupees (about $1.20) per unit – compared to the global market rate of $2 to $3, with government subsidies covering 40 percent of the cost. To manage these efforts efficiently, the government has also launched a centralized AI computing portal, called IndiaAI Compute, to streamline GPU allocation and optimize AI-related infrastructure.
Unlocking Data to Compete in the Global AI Race: India’s AIKosha Initiative
The accessibility of data is essential for competing effectively in the global AI race. The United States has achieved significant success in AI, largely due to the vast amounts of data collected in English, which have supported the training of sophisticated AI models. Similarly, China has rapidly advanced in AI development by leveraging extensive data resources that enable Chinese companies to compete more closely with their U.S. counterparts. Against this backdrop, there has been growing demand from AI startups and policy experts in India for the Indian government to open up and share the substantial data resources it holds.
Responding to this emerging need from India’s private sector, and following the recent development of DeepSeek, the Indian government launched the IndiaAI Dataset Platform, known as AIKosha, on March 6, 2025 – an initiative aimed at establishing a unified platform for comprehensive access to datasets, AI tools, and models. AIKosha addresses the current gaps in India’s AI infrastructure, providing seamless access to high-quality, non-personal datasets. By doing so, it enables Indian academia, researchers, startups, and industries to accelerate innovation and compete more effectively on a global scale.
Looking Ahead: Sustaining India’s AI Ambitions Amid Persistent Challenges
DeepSeek has significantly transformed India’s perception of the AI race, prompting the Indian state to actively reposition itself in global AI competition. Nevertheless, significant challenges remain for India to achieve substantial outcomes in AI development.
Securing a competitive position in global AI innovation requires a sustained, long-term strategy for nurturing high-tech industries. Here, India can draw valuable insights from East Asia’s successful models, particularly from countries like China, South Korea, and Taiwan, which developed their AI and semiconductor sectors through a clear long-term vision, consistent government support, and close collaboration between state institutions and private enterprises. These East Asian nations succeeded by aligning sustained governmental support with private-sector innovation, enabling them to compete against Western tech powers that dominated high-tech industries for decades.
India’s post-DeepSeek initiatives represent a promising start. However, ongoing and sustained collaboration between the government and private sector will be necessary for India to implement a coherent, long-term vision. Only then can India secure a meaningful position in the global AI race.