AI bubble becomes credit investors' biggest fear for the first time

Market sentiment is an inertial thing. When professional investors managing trillions of dollars in debt instruments synchronously shift their main fears, it deserves close attention. This is exactly what happened in February 2026: according to Bank of America's regular survey of credit division clients, "the artificial intelligence bubble" has for the first time in history taken first place among key market risks. Not geopolitics, not inflation, not recession — but rather AI.

To understand the scale of this shift, it's worth recalling the context. Throughout 2023 and 2024, the technology sector experienced euphoria comparable to the dotcom boom. Nvidia's market capitalization grew several times over, Microsoft, Google, Amazon, and Meta increased capital expenditures on data centers and AI infrastructure at rates analysts called unprecedented. In 2025 alone, the combined investments of the "big four" cloud providers in AI infrastructure, by various estimates, exceeded 250 billion dollars. Yet the question of returns on these investments remained open — revenue from AI products was growing, but clearly not at rates that could justify such expenditures.

Credit investors are not venture optimists or retail traders buying stocks on hype. They are bond portfolio managers who assess companies' ability to service debt over horizons of five, ten, twenty years. When exactly this category of market participants begins calling the AI bubble their chief concern, we are talking not about abstract philosophical discussions, but about concrete credit spreads, borrowing costs, and capital availability for the entire technology sector. Essentially, the debt market is sending a signal: we are no longer confident that colossal AI spending will pay off, and we are prepared to price this risk in.

There have been enough reasons to accumulate such pessimism. First, revenue growth rates from generative AI at most large companies began slowing toward the end of 2025 — not because the technology is poor, but because corporate clients proved slower at adoption than expected. Implementing AI in business processes requires not only API subscriptions, but infrastructure overhaul, staff training, data security decisions. Second, competition between AI model providers led to a sharp drop in inference prices, which is good for consumers but creates pressure on margins. Third, the emergence of effective open-source models has called into question the very business model under which companies spend billions training proprietary systems.

It is important to understand that credit investors' concerns are not a verdict on the technology. The internet bubble of the early 2000s burst, but the internet didn't go anywhere and ultimately transformed the economy exactly as the boldest visionaries promised. The problem was not the technology, but valuations and monetization timelines. The analogy to AI suggests itself: artificial intelligence truly is transforming entire industries, but the path from laboratory demonstrations to sustainable corporate revenue turned out to be longer and more winding than markets had factored in during the euphoria of 2023-2024.

The practical consequences of this shift in sentiment could be quite tangible. If credit markets begin systematically repricing risks in the technology sector, borrowing costs will rise across the entire chain — from the largest cloud providers financing data center construction to AI startups attracting venture debt financing. This will not necessarily lead to a collapse, but could significantly slow the pace of infrastructure investment. For AI startups, the signal is even more concerning: in conditions where even credit investors doubt the sustainability of the trend, securing next-round financing will become significantly harder.

Bank of America's survey captures not a crash, but a turning point in perception. The market is transitioning from a phase of unconditional faith in AI to a phase of demanding verification: show us revenue, show us returns on investment, show us that your clients truly are paying, not just experimenting. This is a painful, but ultimately healthy process.

Companies with real products and proven value for clients will make it through this period of doubt. Those who built their business on presentations and promises will find themselves in a far more difficult position. The era of free money for anything containing the letters "AI" in its name is coming to an end — and this may well be the best thing that could have happened to the industry.