AI Investment Bubble Could Trigger Global Shock Worse Than 2008 Financial Crisis, Says Economic Survey

The Economic Survey 2025–26 has cautioned that the world could face a financial shock in the coming years, which could leave scars deeper than those of the 2008 global financial crisis. In a worst-case scenario, a “systemic shock” triggered by an artificial-intelligence (AI) investment bust, coupled with “geopolitical escalation or trade disruption”, could wreak havoc across the financial landscape, with far greater intensity and magnitude than the 2008 global financial crisis.

“The recent phase of highly leveraged AI-infrastructure investment has exposed business models that are dependent on optimistic execution timelines, narrow customer concentration, and long duration capital commitments,” it said. Such a scenario, it added, would not end technological adoption, but could tighten financial conditions, trigger risk aversion and spill over into broader capital markets. If that were to happen in tandem with escalation in geopolitical tensions or trade disruption, it could lead to sharper contraction in liquidity, sudden weakening of capital flows, which, in turn, could push countries into adopting defensive economic policies.

The survey said the probability of such an occurrence was 10-20 per cent. Nonetheless, the consequences could be “significantly asymmetric.” At the heart of the threat was the boom in AI-heavy infrastructure, which is built on long-term funding, optimistic project schedules, and reliance on just a few major customers, it added.

Implications For India's IT Sector Jobs

For India, the implications could extend beyond markets to employment. The Survey highlighted growing stress in the country’s IT and services sector, long seen as the engine of white-collar job creation. Using trends in the US Professional, Business and Information Services sector as a reference, it observed a clear shift after late 2022, when generative AI went mainstream. Since then, economic growth has no longer translated into proportionate job creation in these segments, it added.

According to the survey, jobs have not disappeared overnight, but the relationship between output growth and hiring has weakened, particularly for roles involving information processing, routine analysis, and standardised cognitive tasks. This change was gradual but led to a persistent drift in labour demand, which is largely confined to highly digitised sectors. Traditional industries, however, showed no such comparable break.