Beyond an application layer

The global artificial intelligence (AI) landscape is witnessing a significant transformation. Even as the US and China remain the two countries specialising in distinct but complementary segments of the AI value chain, other countries are now engaged in a fierce scaling up with India increasingly working towards that objective.

The US remains the undisputed leader in frontier foundation models, driven by companies such as OpenAI, Anthropic, Google DeepMind, Meta, and xAI, supported by unparalleled venture capital, world-class research universities, and access to cutting-edge computing infrastructure. The US produces the largest share of frontier AI models and continues to lead in breakthrough innovation through advanced large language models, reasoning models, and agentic AI systems.

In contrast, instead of competing exclusively on frontier model development, China has carved out a distinct comparative advantage in industrial AI and robotics. China has successfully integrated AI into manufacturing, logistics, autonomous systems, and industrial automation through coordinated industrial policy and its manufacturing ecosystem. China now leads the world in industrial robot installations, AI patents, and AI-related scientific publications, while Chinese firms have rapidly closed the performance gap in foundation models through open-source development and large-scale deployment. More importantly, Chinese companies have excelled at embedding AI into the real economy — making AI a productivity-enhancing technology rather than simply a digital service.

Where does India stand at this juncture? India is the only country to have successfully organised an AI Conclave with almost all the competing players on one single platform. Simultaneously, rather than developing new foundation models or breakthrough AI architectures, India is taking a different path by increasingly focusing on adapting existing technologies to local needs by using the application layer it has so effectively built up in the last decade. Thus, India is emphasising small language models, multilingual applications, digital public infrastructure, and domain-specific deployment in healthcare, education, finance, and governance.

However, the question is, should India remain merely the world’s application layer?

While an application-first strategy is economically sensible in the short run given India’s resource constraints, remaining confined to this layer in the long run will entail technological dependence on the US and China. In effect, India risks recreating its historical position in the global IT industry: consumer and integrator of AI innovations rather than a creator of these by remaining a leader in implementation and services, but not in ownership of the core technologies that define the technological frontier.

If India aspires to become more than an application-layer economy, it must first develop a realistic assessment of its position in the global AI ecosystem. India’s comparative advantage today lies in its digital public infrastructure, software engineering talent, and AI applications rather than frontier innovation. Instead of attempting to compete simultaneously across every dimension of AI, India should identify those frontier capabilities where it can realistically build comparative advantage over the next decade.

We could do three things.

Firstly, we need two complementary national strategies: a short-term road map (two-three years) focused on rapidly strengthening AI infrastructure, talent, and institutional capacity, and a long-term road map (10 years) aimed at positioning India as a creator of frontier AI technologies rather than merely an adopter. To formulate these strategies, the government should constitute two independent yet complementary committees. The first should be a parliamentary committee to build bipartisan political consensus and ensure continuity across electoral cycles. The second should be an expert commission comprising economists, computer scientists, AI researchers, industry leaders, national security experts, and public policy scholars. AI is not merely a technological issue; it is a general-purpose technology that will reshape labour markets, productivity, competition, inequality, public finance, education, healthcare, and national security.

Secondly, India should move beyond relying exclusively on its existing engineering institutions. While the IITs and IISc remain world-class educational institutions, the institutional structures, compensation systems, and research incentives may be further designed to compete for the world’s top AI researchers against organisations such as OpenAI, Google DeepMind, Anthropic, or leading American universities. India could therefore establish an Indian Institute of Artificial Intelligence (IIAI) as a globally competitive research institution with even equity stakes from IITs/IISc as well as the government of India with operational autonomy and internationally benchmarked salaries. The objective should be to attract leading Indian and international AI scientists, much as institutions such as the King Abdullah University of Science and Technology in Saudi Arabia, the Mohamed bin Zayed University of Artificial Intelligence in the UAE, or the Canadian Institute for Advanced Research in Canada have successfully done. IIAI should establish joint research programmes, dual faculty appointments, and long-term collaborative laboratories with such leading global AI institutions.

Thirdly, the creation of a robust AI research ecosystem through large-scale public compute infrastructure, access to high-quality data sets, and international collaboration. Appropriate frameworks should enable privacy-preserving data-sharing between government, academia, and large private firms while safeguarding commercial confidentiality and individual privacy.

India’s IT industry became a global powerhouse by excelling in software services, systems integration, and implementation without owning the underlying computing platforms. However, unlike traditional software, value in AI increasingly accrues to those controlling the foundational models, computing infrastructure, semiconductor ecosystem, and robotics platforms. And in the long run, India should start investing in such technologies (as is being done in semiconductors across India) and emerge as a third global AI powerhouse.

The writers are respectively Member of 16th Finance Commission and EAC-PM & Group Chief Economic Advisor, SBI; and Member of 8th Pay Commission and EAC-PM & Professor, IIM Bangalore.

Disclaimer: The views expressed are the author’s own and do not reflect the official policy or position of Financial Express.