AI Valuations Spark Bubble Fears, CEOs Say Demand Will Hold

Executives from leading AI companies acknowledged a growing unease about valuation levels even as they defended the longer term promise of the technology. The CEOs of DeepL and Picsart told CNBC they were concerned about overstretched AI valuations, while emphasizing that businesses are already finding productivity gains and new uses for generative models.

"I think there's a lot of demand, and there's a lot of interest. I think everybody understands that AI can do magical things to businesses, and... we can all operate on another level when it comes to efficiency," Kutylowski said, characterizing the optimism that is fueling investor interest even as some question whether that enthusiasm is translating into durable revenue.

Investors and analysts are debating whether the market is experiencing a classic speculative bubble, with some warning that a correction could be imminent. That debate is taking place against a backdrop of enormous capital commitments to infrastructure and compute. A report from venture capital group Accel released this week forecasts that the buildout of new AI data center capacity will reach 117 gigawatts by 2030. Accel translates that capacity need into about four trillion dollars of capital expenditure over the next five years.

Those numbers suggest two simultaneous dynamics at work. On the demand side, corporate customers are increasing spending on AI tools and services, with executives telling CNBC they expect enterprise demand to remain strong into 2026. On the supply side, investors and cloud providers are committing to a large scale of compute and real estate expansion. The combination raises the risk of a disconnect where valuations race ahead of near term revenue, while heavy upfront investment locks in capacity that may outstrip demand growth.

The market implications are immediate. Public cloud providers and chipmakers are likely to see persistent orders and backlog for specialized hardware and data center services, supporting revenue for several years even if software companies face a valuation reset. Venture funded startups could face a squeeze if private market valuations snap back, making follow on financing more expensive and consolidations more likely. For investors, timing a reallocation between platform infrastructure and application layer software will be critical.

Policy makers and regulators are watching the sector closely as well. Large capex projects raise questions about regional power grids, permitting and environmental impacts, and workforce needs for high skilled operations. If a correction in valuations reduces investment momentum, regions that have positioned themselves as AI hubs may confront stranded assets and employment disruptions.

Long term, the outlook hinges on whether enterprise AI adoption converts curiosity and pilot deployments into sustained spending. If efficiency gains materialize at scale across industries, revenues could catch up to valuations over several years. If much of the market is driven by what some observers call vibe revenue, meaning revenue powered by hype rather than durable product market fit, then a period of consolidation and price discovery may follow.

For now the paradox persists. Capital is flowing at a scale not seen in previous tech cycles, even as public and private markets wrestle with whether current price tags reflect the real economics of AI adoption.