AI project failure rates surge, fueling tech stock bubble fears
Surging abandonment rates for artificial intelligence initiatives are prompting analysts to warn of an imminent correction in overvalued tech stocks.
Corporate America is rapidly scaling back its artificial intelligence ambitions. An S&P Global executive survey revealed that 42 per cent of organisations abandoned most of their AI initiatives in 2025, a sharp increase from 17 per cent the previous year. This retreat is mirrored in the industrial sector, where a 2024 RAND report found that more than 80 per cent of AI projects fail due to process complexity and poor data quality.
These operational setbacks are fueling concerns that a market bubble is forming around the technology. Investors have poured capital into the S&P 500 and Nasdaq, largely rewarding a handful of tech giants including Nvidia, Microsoft, and Apple. However, veteran investor Jeremy Grantham plans to sell his tech holdings, warning that AI will follow the historical pattern of railways and the internet. Once the market realises AI is merely a utility like electricity, Grantham argues, the windfall profits expected by early investors will fail to materialise.
The industrial sector offers a clear window into why these deployments are stumbling. While traditional automation succeeds in stable environments, manufacturing plants are inherently volatile. Late supplier deliveries, equipment breakdowns, and shifting regulatory constraints require adaptive problem-solving that current AI systems cannot handle. When companies attempt to apply AI to these variable operations, they risk wasting capital on technology incapable of navigating real-world friction.
Ford’s recent experience illustrates the dangers of over-reliance on immature AI. The automaker integrated AI to speed up decision-making and vehicle development, only to discover the systems buckled when fed incomplete data. As veteran engineers departed, critical institutional knowledge vanished from the training datasets. "Mistakenly, we thought that by just introducing artificial intelligence and adjusting the design requirements that we had, that that would produce a high-quality product," said Charles Poon, Ford’s vice president of vehicle hardware engineering. The company has since reversed course, rehiring over 350 experienced engineers to improve data collection.
Despite these setbacks, AI is not without industrial value when applied correctly. As a January Forbes article noted, AI must operate within existing production systems and real workflows, with humans remaining accountable for the final outcome. For equity markets, the implication is clear: AI will likely enhance productivity and enable predictive maintenance, but it will not replace human judgment. Investors banking on a seamless AI revolution may need to adjust their valuation models for a much slower, more incremental adoption curve.
