Jefferies warns AI investment cycle may have peaked

The peak of the AI boom

The global investment cycle for artificial intelligence may be reaching a turning point. According to a report from Jefferies, the massive spending on AI infrastructure could be peaking or entering a period of correction. This shift is driven by growing concerns over whether the current level of spending will actually lead to sustainable profits.

Financial markets are beginning to show signs of hesitation regarding the funding of new AI projects. While the technology has seen massive investment, the sheer scale of capital required is starting to create friction. Investors are increasingly looking for clear evidence of returns before committing more capital to the AI supply chain.

The report suggests that the era of unchecked spending might be cooling down. As the costs of financing these projects rise, the sustainability of the current spending habits by major US technology companies is coming under scrutiny. This could lead to a period where investment growth slows down significantly compared to previous years.

Hyperscaler spending trends

The scale of current spending is immense, particularly among the four major hyperscalers: Microsoft, Amazon, Alphabet, and Meta. These companies grew their total spending on capital projects by 108 percent in a single year, reaching USD 202 billion during the third quarter of 2026. This massive outlay highlights just how much capital is currently flowing into the AI ecosystem.

Despite this high spending, the outlook for 2027 remains a point of contention. Consensus estimates currently place the combined capital expenditure for these four companies at USD 1.04 trillion for the upcoming year. However, Jefferies warns that these figures might not be met if concerns over profitability and financing continue to weigh on the market.

The report notes that AI-related investment has already had a significant impact on the economy. It contributed 1.39 percentage points, or 67 per cent, to the US real GDP growth of 2.1 per cent during the four quarters ending in the third quarter of 2026. This demonstrates the massive scale of the current cycle even as it faces potential headwinds.

Rising costs and credit risks

One of the primary drivers for a potential slowdown is the rising cost of borrowing. Technology companies are facing wider bond yield spreads, which indicates that it is becoming more expensive for them to finance their operations. This credit-driven pressure could force a reevaluation of how quickly companies can scale their AI ambitions.

Specific companies are already showing signs of increased credit risk. The report highlights that credit default swap spreads for Oracle and CoreWeave have increased sharply. These metrics are used to gauge the likelihood of a company defaulting on its debt, and their rise suggests a more cautious environment for tech lending.

A slowdown in AI spending due to these credit issues could have broader economic consequences. It might negatively affect US economic growth and could even trigger a reversal in Federal Reserve policy. Instead of continuing with monetary tightening, the Fed might move toward easing if the AI-driven spending engine loses steam.

Shifting investment targets

As the AI supply chain faces scrutiny, Jefferies is advising investors to look elsewhere for growth. Specifically, the firm suggests exploring opportunities in healthcare and biotechnology. They note that AI applications in drug development offer significant potential that may be less volatile than the core AI infrastructure market.

The firm is also adjusting its regional asset allocations to reflect these changing priorities. Jefferies increased the weight of India within its Asia-Pacific ex-Japan relative-return portfolio by two percentage points. This move brings the allocation to 14 per cent, maintaining an Overweight stance on Indian equities.

Conversely, the firm reduced its allocations to South Korea and Taiwan by one percentage point each. These markets remain in Underweight positions. These adjustments reflect a broader strategy to find value in different regions and sectors as global investors reassess the primary themes driving the technology market.

The road ahead for AI

The transition from a peak to a correction phase does not mean that AI development will stop. Rather, it suggests a move toward a more disciplined approach to spending. Companies will likely need to prove that their investments are generating tangible value to justify the high costs of operation.

The report suggests the AI capex cycle may have peaked, according to Jefferies. This sentiment underscores a shift from pure expansion to a focus on the sustainability of the current technological trajectory.

For the broader market, this means a period of increased volatility as investors weigh the potential of AI against the reality of high interest rates. The next few quarters will be critical in determining whether the AI boom can transition into a period of stable, profitable growth or if it will face a significant contraction.