The Economic Survey 2025–26 has issued a rare and pointed warning on the financial risks embedded in the global artificial intelligence (AI) boom, cautioning that highly leveraged investments in capital-intensive AI infrastructure could emerge as a new source of systemic instability, with spillovers extending well beyond the technology sector.
Tabled in Parliament by Union minister of finance and corporate affairs Nirmala Sitharaman, the Survey frames AI not merely as a transformative technology but as a macroeconomic and financial variable with the potential to amplify risks across markets, investors and institutions. In doing so, it marks one of the clearest acknowledgements yet by the Union government that the next global financial shock may originate from technology finance rather than traditional banking channels alone.
The Survey notes that the global AI race is increasingly being driven by massive capital commitments to compute-heavy frontier models, data centres and specialised hardware. These investments are often funded through layered financial structures involving private equity, venture capital, debt instruments and public markets, creating channels through which stress could propagate rapidly.
“Highly leveraged investments in AI infrastructure could trigger cascading financial shocks,” the Survey cautions, underlining that exposure is not confined to technology companies. "Pension funds, retail investors, insurance pools and broader capital markets are increasingly tied to the valuation and performance of AI-driven enterprises."
"This concentration of capital, combined with uncertain revenue models and long gestation periods, raises the risk of sharp corrections if expectations around AI adoption, productivity gains or monetisation fail to materialise at the pace currently priced in by markets," the Survey says.
Against this backdrop, the Survey articulates a conscious policy choice for India: to avoid replicating the capital-intensive frontier AI model pursued by advanced economies and instead prioritise decentralised, application-driven systems aligned with domestic economic realities.
India’s approach, the Survey stresses, is grounded in constraints related to capital availability, energy intensity, institutional capacity and market depth. "Rather than building a few large, centralised AI systems that could create fragile dependencies, the preferred pathway is one of smaller, task-specific models deployed across sectors."
Such systems, the Survey argues, diffuse innovation more evenly, reduce entry barriers for firms, and are better suited to India’s diverse and decentralised economy.
Unlike speculative AI deployment focused on scale or prestige, India’s demand for AI is emerging from practical, real-world problems. The Survey points to growing adoption across healthcare, agriculture, urban governance, education, disaster management and public administration.
Examples include early disease screening, precision irrigation, farmer market access platforms, classroom analytics and regional language interfaces. These applications typically operate on local hardware, function in low-resource environments and prioritise cost reduction and efficiency over computational scale.
This signals, the Survey notes, a large and scalable market for frugal, application-focused AI solutions—one that generates economic value without the systemic risks associated with heavy leverage and capital concentration.
A recurring theme in the Survey is the role of open and interoperable AI systems as a counterbalance to financial and technological fragility. It says, decentralised models reduce dependence on a narrow set of global providers for compute, capital and standards, thereby insulating domestic systems from external shocks.
The national AI mission is positioned as a key enabler in this ecosystem, providing shared infrastructure, governance frameworks, standards and funding, while preserving space for local creativity. Municipal bodies, start-ups, community institutions and local innovators are identified as critical actors in deploying AI solutions tailored to contextual challenges.
Language and voice-first AI systems are also highlighted as extending the reach of digital services to populations historically excluded from formal digital ecosystems, reinforcing the Survey’s emphasis on inclusion over scale.
The Survey also reframes the skills debate in the AI era, warning against over-investment in narrow technical specialisation that may become obsolete as technologies evolve. Instead, it advocates strengthening foundational human capabilities such as reasoning, communication, judgement and adaptability.
Education and training systems, the Survey argues, must integrate AI into workplaces and public systems while preserving human agency. Initiatives such as experiential learning, flexible education pathways and 'earn and learn' models are identified as essential to aligning AI adoption with labour market realities.
On governance, the Survey calls for proportionate, risk-based regulation rather than rigid controls. Data governance, it argues, should focus on accountability and value creation rather than isolation, enabling trusted data flows while ensuring that economic value from domestic data accrues within India.
Importantly, the Survey warns against premature regulatory lock-in that could entrench fragile models or stifle innovation. Instead, it recommends sequencing coordination first, capacity next and binding policy last, allowing institutions and markets to co-evolve.
In the context of financial stability, this approach implicitly recognises that unchecked AI-led capital cycles—similar to past episodes in real estate or structured finance—could amplify volatility if governance lags behind innovation.
The Survey supports the establishment of an AI Safety Institute under the Union ministry of electronics and information technology (MeitY) to assess emerging risks, conduct scenario-based testing and coordinate on safety standards. It also calls for international cooperation with institutions such as the UK’s AI Security Institute and the US National Institute of Standards and Technology to jointly evaluate high-risk models.
Such coordination, the Survey notes, is essential in a world where AI systems and their financial backers operate across borders, making isolated national safeguards insufficient.
The Economic Survey ultimately positions India’s AI strategy as a calibrated response to a rapidly evolving global environment, one that balances opportunity with caution. While acknowledging AI’s transformative potential, it underscores that technology-led growth must not come at the cost of financial fragility.
With global markets increasingly exposed to leveraged bets on AI infrastructure, the Survey’s warning stands out as a rare official acknowledgement that systemic risk may be building outside the traditional banking system.
India’s task, it concludes, is to harness AI for productivity, inclusion and dignified employment while avoiding the excesses that have historically preceded financial crises—ensuring that innovation strengthens, rather than destabilises, long-term economic resilience.
