Artificial intelligence has become the biggest source of concern for global fund managers, with 65% identifying either a potential AI bubble or a credit event involving major AI infrastructure companies as the greatest risk to financial markets. The findings come from Bank of America’s November 2026 Global Fund Manager Survey, which points to a sharp change in investor sentiment as spending on artificial intelligence continues to expand.

The combined share of fund managers worried about these two AI-related risks nearly doubled from 34% in October to 65% in November. The shift suggests that investors are increasingly questioning whether the enormous investments being made in data centres, chips and cloud infrastructure will generate sufficient returns—and whether the debt used to finance this expansion could create broader financial instability.

Key Takeaways

  • 65% of fund managers identify an AI bubble or a potential hyperscaler credit event as the biggest market tail risk, compared with 34% in October.
  • 34% are concerned about an AI bubble, up from 23% in the previous month.
  • 31% identify a potential credit event involving AI hyperscalers, compared with 11% in October.
  • 52% expect at least one AI hyperscaler to reduce capital expenditure over the next six months.
  • A record net 38% believe companies are overinvesting, while just 16% want corporations to increase capital expenditure.
  • Investors remain net overweight global equities, but have significantly reduced their exposure to technology stocks.

AI Bubble and Hyperscaler Debt Become Major Market Risks

The November survey highlights two closely connected concerns: the possibility that valuations across the artificial intelligence industry have become excessive and the risk that borrowing by major technology companies could create a credit event.

The share of fund managers identifying an AI bubble as the biggest tail risk increased to 34% in November from 23% in October. Meanwhile, concerns about a potential credit event involving AI hyperscalers climbed to 31% from 11%.

Together, these risks account for almost two-thirds of the survey’s responses, making AI-related concerns substantially more prominent than traditional market worries such as inflation and rising bond yields.

Biggest market tail riskOctober 2026November 2026
AI bubble23%34%
AI hyperscaler credit event11%31%
Combined AI-related risks34%65%

The figures represent the share of respondents identifying each issue as the biggest tail risk. They reflect investor perceptions, not the probability that a market crisis will occur.

The change is significant because investors are no longer focusing exclusively on whether AI-related stocks are overvalued. They are also examining how the infrastructure expansion is being financed and whether the companies making these investments can generate enough cash flow to support their spending.

Why Investors Are Questioning the AI Investment Boom

The AI industry has entered an exceptionally capital-intensive phase. Building and operating large-scale AI systems requires substantial investment in data centres, advanced processors, networking equipment, electricity infrastructure and cooling systems.

Major cloud providers and technology companies are committing enormous resources to expand their computing capacity. These investments are intended to support demand for AI models, enterprise applications and cloud-based AI services.

However, the financial returns from this infrastructure are not guaranteed to arrive at the same pace as the spending.

Companies must invest in facilities and equipment before they can fully establish how much revenue those assets will generate. If demand grows more slowly than expected, or customers are unwilling to pay enough for AI services, the returns on these investments could fall short of expectations.

This creates a potential mismatch between the capital being deployed today and the revenue that may materialise over several years.

Fund managers appear increasingly concerned about this possibility. The survey found that a record net 38% of respondents believe companies are overinvesting, while only 16% want corporations to increase capital expenditure.

The findings suggest that investors are becoming more selective about the scale and financial discipline of AI infrastructure spending.

52% Expect AI Hyperscalers to Cut Capital Expenditure

One of the strongest signals in the survey concerns expectations for future investment.

Around 52% of fund managers expect at least one AI hyperscaler to announce a reduction in capital expenditure over the next six months. This represents a substantial increase from 30% in October.

Capital expenditure, commonly known as capex, refers to money companies spend on long-term assets such as buildings, data centres, servers, networking equipment and other infrastructure.

For AI hyperscalers, these investments are central to expanding computing capacity and competing in the artificial intelligence market.

However, a reduction in spending would not necessarily mean that companies are abandoning AI. It could indicate that management teams are adjusting investment plans, prioritising projects with clearer returns or responding to concerns about financing costs.

The survey also shows a growing difference between investor expectations and corporate investment ambitions. While AI companies have been competing to build infrastructure, a significant proportion of fund managers now want them to demonstrate stronger financial discipline.

What a capex slowdown could mean

A reduction in AI infrastructure spending could have several consequences:

  1. Data centre development: Slower investment could affect the pace of new data centre construction and expansion.
  2. Semiconductor demand: Companies supplying AI processors and related equipment could face changes in future orders if infrastructure spending moderates.
  3. Cloud services: Hyperscalers may focus more heavily on utilisation rates and the profitability of existing infrastructure.
  4. AI startups: Smaller companies that depend on access to large computing platforms could face changes in pricing, availability or investment conditions.
  5. Technology valuations: Investors could reassess companies whose market valuations depend heavily on expectations of uninterrupted AI spending growth.

These are potential consequences rather than confirmed outcomes. The survey measures fund managers’ expectations, not announced spending decisions by individual companies.

AI Hyperscaler Bonds Emerge as a Credit Risk

The concerns extend beyond stock market valuations to the debt financing behind the AI expansion.

When asked where a systemic credit event was most likely to originate, 38% of respondents pointed to AI hyperscaler bonds. Private credit was identified by 30%, while government debt was cited by 15%.

Potential source of a systemic credit eventShare of respondents
AI hyperscaler bonds38%
Private credit30%
Government debt15%

The responses indicate that fund managers are paying greater attention to how large technology companies finance their infrastructure plans.

A credit event could emerge if borrowers struggle to meet their financial obligations, refinancing becomes more expensive or investors lose confidence in the ability of companies to generate sufficient cash flows.

The risk is particularly relevant for capital-intensive AI infrastructure projects. Data centres and computing facilities require large upfront investments, while their financial returns depend on future demand, pricing and utilisation.

If revenue growth disappoints, companies with substantial debt obligations could face pressure. A broader reassessment of the creditworthiness of AI infrastructure borrowers could also affect bond investors and lenders.

However, identifying hyperscaler bonds as a potential source of systemic risk does not mean that a default is imminent. It reflects the concern that growing borrowing and concentrated exposure to the AI investment cycle could create vulnerabilities if market conditions deteriorate.

Fund Managers Are Reducing Technology Exposure

The change in sentiment is already visible in portfolio positioning.

According to the survey, fund managers reduced their overweight allocation to technology stocks by the largest amount since June 2006. This indicates a meaningful shift away from the sector after a period in which AI-related optimism supported strong investor interest.

Long positions in global semiconductor stocks remained the most crowded trade, cited by 39% of respondents. A crowded trade is an investment position held by a large number of investors, which can make market movements more pronounced if sentiment changes and many investors attempt to exit simultaneously.

Other crowded positions included short China stocks, cited by 14%, and long positions in the Magnificent Seven stocks, cited by 12%.

The reduction in technology exposure does not necessarily mean fund managers expect the entire AI industry to decline. Rather, investors may be becoming more cautious about valuations, competitive pressures and the ability of companies to convert infrastructure spending into sustainable profits.

For technology companies, this could mean that investors increasingly demand evidence of revenue growth, operating margins and returns on invested capital instead of relying primarily on long-term AI growth projections.

Investors Remain Positive on Global Equities

Despite the growing concerns around AI, the survey does not indicate a broad withdrawal from financial markets.

Average cash allocations among fund managers declined to 3.6% of assets under management in November from 3.9% in October. Investors also remained net overweight global equities.

This distinction is important. Fund managers can become cautious about a particular sector without turning negative on the broader stock market.

The survey found that half of respondents expect a soft landing for the global economy over the next 12 months. Another 42% expect a no-landing scenario, in which economic growth remains above trend. Only 3% anticipate a hard landing.

Meanwhile, traditional market concerns have become less prominent.

Inflation was identified as the biggest tail risk by 14% of respondents, while 13% pointed to a disorderly rise in bond yields. Both concerns had attracted greater attention in the previous month’s survey.

The results suggest that investors are shifting their attention towards risks associated with the AI investment cycle rather than anticipating an immediate global economic downturn.

What the Survey Means for Semiconductor and Technology Companies

The findings could have implications across the technology supply chain.

Semiconductor manufacturers, data centre operators, cloud providers, networking equipment suppliers and electricity infrastructure companies have benefited from expectations of sustained AI investment.

However, if hyperscalers moderate their capital expenditure, investors may begin examining which companies are most exposed to changes in infrastructure demand.

Chip manufacturers could face questions about future order growth. Data centre operators may need to demonstrate that new capacity can be leased and used profitably. Cloud providers could face pressure to show that AI services are generating enough revenue to justify their infrastructure investments.

The survey also highlights the importance of distinguishing between long-term AI adoption and short-term investment expectations.

Artificial intelligence could continue expanding across industries even if infrastructure spending slows temporarily. Companies may still adopt AI tools for software development, customer service, research, automation and enterprise productivity.

The central question for investors is whether the financial value generated by these applications will justify the capital being deployed to build the underlying infrastructure.

What the Findings Could Mean for Indian Markets

Although the survey focuses on global fund managers, its findings are relevant to Indian technology and financial markets.

Indian IT services companies are increasingly investing in AI capabilities and integrating generative AI into software development, consulting and enterprise services. A more cautious global investment environment could influence technology budgets, client spending priorities and demand for AI-related services.

Indian companies supplying IT services, engineering solutions and data centre infrastructure could also be affected by changes in global technology investment plans.

The semiconductor ecosystem is another area to watch. India’s ambitions in electronics manufacturing, semiconductor fabrication and data centre development are connected to the broader global technology investment cycle.

If major hyperscalers slow their infrastructure expansion, the timing and scale of some investment opportunities could change. However, this would not automatically undermine India’s longer-term digital infrastructure ambitions.

For Indian investors, the survey is a reminder that global technology valuations and international capital allocation decisions can influence market sentiment even when the underlying businesses operate in different countries.

The Bigger Picture

The November BofA survey captures an important transition in the global AI investment narrative. Investors are moving beyond enthusiasm about the technology’s potential and asking more difficult questions about the economics of the infrastructure supporting it.

The scale of investment in AI has created opportunities for semiconductor manufacturers, cloud providers, data centre operators and technology companies. But it has also increased the importance of capital discipline, financing costs, utilisation rates and the ability to generate cash flows.

The growing concern about hyperscaler bonds is particularly significant because it connects the AI investment cycle with the broader credit market. If infrastructure spending is increasingly financed through borrowing, disappointing returns could affect not only technology valuations but also lenders and bond investors.

At the same time, the survey does not establish that an AI bubble has already formed or that a credit crisis is inevitable. It shows that a growing share of professional investors considers these risks important enough to rank them above traditional market concerns.

Looking Ahead

The next six months could provide an important test of investor expectations. Fund managers will be watching whether major AI hyperscalers maintain their infrastructure budgets or begin moderating capital expenditure. Their decisions, alongside updates on AI revenue growth, data centre utilisation and financing requirements, could influence valuations across the global technology sector.

For investors, the central issue is likely to shift from how quickly artificial intelligence can expand to how profitably companies can build and operate it. AI adoption may continue growing, but the market could become less forgiving of businesses that commit large amounts of capital without demonstrating sustainable financial returns. The November survey suggests that investors are increasingly demanding that distinction.

Get the day’s top stories in your inbox

One concise email. No spam, unsubscribe anytime.