The Economic Survey 2025–26 has outlined a sweeping reorientation of India’s artificial intelligence strategy, signalling the government’s intent to treat AI as a public good through a mix of shared infrastructure, decentralised innovation, education reform and labour-market redesign, while also enabling customised AI tools tailored to personal, individual, group, organisational and local needs across sectors and regions.
In a marked departure from dominant global trends centred on frontier-scale large language models, the Survey has proposed the creation of an ‘AI-OS’ initiative, under which the government would act as a monetary shareholder in core AI infrastructure, much like its role in building public digital goods such as Aadhaar and the Unified Payments Interface.
The proposed framework envisions artificial intelligence as a common digital backbone, with shared access to compute, standards and governance, while allowing innovation to emerge organically from diverse actors.
At the heart of the proposal is a centralised code repository under the IndiaAI mission, designed to function as a government-supported equivalent of GitHub, where developers, researchers, start-ups, enterprises and public agencies can collaboratively build, share and refine AI tools.
By providing common infrastructure, funding and oversight, the state aims to lower entry barriers and enable distributed innovation without constraining local creativity or sectoral specificity.
Rather than expending scarce capital on building and training expensive frontier models, the Survey has argued that India should prioritise application-specific, smaller AI models optimised for clearly defined sectoral needs.
These computationally efficient systems, capable of running on widely available hardware such as smartphones and personal computers, are seen as better suited to India’s resource constraints, as they reduce dependence on large data centres while allowing AI adoption to spread across firms, research institutions and public services.
The Survey has framed this as a “bottom-up” pathway to AI development, one that contrasts with the capital-intensive and energy-heavy approach pursued in many Western economies, and it has noted that India’s relatively late entry into the AI race may, paradoxically, offer strategic advantages.
Citing international benchmarks, it has pointed out that India already possesses one of the world’s most AI-literate workforces, ranking second globally, which strengthens the feasibility of decentralised and sector-driven innovation.
A major pillar of the strategy is a proposed restructuring of school and higher education through an ‘Earn-and-Learn’ model, under which students could begin accumulating academic credits alongside paid work experience from Class 11 onwards.
Industry-designed, credit-bearing apprenticeships and fellowships are envisaged as a way to integrate formal education with real-world projects, thereby aligning skill formation with the emerging AI economy and unlocking the country’s demographic potential.
Beyond the technology sector, the Survey has emphasised the need to identify high-skill, non-white-collar occupations that remain understaffed, particularly in areas such as nursing, geriatric care, advanced trades, healthcare, hospitality design and education.
By mapping these sectors and upgrading training infrastructure, policymakers aim to mitigate potential job displacement from AI while creating new avenues for meaningful employment.
At the same time, the Survey has sounded a note of caution, highlighting India’s limited access to cutting-edge compute, its small share of global training data, and its vulnerability to concentrated global GPU supply chains, while also warning that slow adaptation could erode the traditional strengths of the IT sector.
It has concluded that India’s AI journey will require careful trade-offs between scale and efficiency, openness and stewardship, innovation and regulation, and strategic autonomy and global integration, as the country seeks to embed AI deeply into its economy without overstretching its resources.