Over the last two years, artificial intelligence (AI) has evolved from a promising technology trend to a corporate survival strategy. CEOs mention AI on nearly every earnings call. Boards demand “AI transformation plans.” Venture capital pours into anything with “AI” in the pitch deck. Yet beneath the headlines and optimism for true disruption lies a growing disconnect: much of today’s enterprise AI investment activity is not being driven by measurable return on investment (ROI), but by fear, uncertainty, and doubt (FUD). More specifically, companies are afraid of being left behind – the classic Fear of Missing Out (FOMO).
Recent evidence increasingly supports this view. An IBM study1 found that CEOs expect AI investments to more than double over the next two years, even while many acknowledge serious operational hurdles and fragmented technology adoption. Notably, 50% of CEOs surveyed admitted that rapid AI investment has already created disconnected systems across their organizations. Historically, technology investments were expected to rapidly improve operational efficiency, reduce costs, and/or create new revenue streams. Many organizations appear to beinvesting first and asking business questions later (or not at all).
Why? Because nobody wants to be the company that “missed AI,” and corporate leaders fear being perceived as technologically obsolete.
This fear-driven behavior is especially visible in venture capital and public markets. According to PitchBook2, AI startups captured nearly 58% of all global venture dollars in early 2025. Investors themselves openly described the environment as a “fear of somebody else winning your market.” That’s not disciplined investing; it is reactive urgency and speculation.
The same psychology is influencing enterprises. Companies feel pressure from investors, analysts, customers, and competitors to demonstrate that they have a credible AI roadmap. But in many cases, these roadmaps lack operational maturity, measurable outcomes, or integration with the broader business strategy. The outcome? Organizations are launching pilots, proofs-of-concept, and internal AI tools without clear economic justification.
The consequences are becoming visible.
A recent report on AI cost management3 found that 80% of enterprises miss their AI infrastructure forecasts by more than 25%, while 84% report meaningful erosion of gross margins due to AI workloads. In other words, companies do not fully understand what they are spending, nor whether the spending is producing business value.
Meanwhile, many AI projects never progress beyond experimentation. TechRadar recently reported4 that 74% of companies have already rolled back or shut down at least one AI customer service initiative because of governance and operational failures. This mirrors what many enterprise technologists quietly admit behind closed doors: AI demos are easy; enterprise-grade deployment is hard.
Even among organizations aggressively promoting AI adoption, there are signs of skepticism. Goldman Sachs analysts recently warned5 that portions of current AI spending appear “unprecedented and unsustainable,” particularly where infrastructure investments are not producing corresponding earnings growth.
So what to do?
The deeper problem is that most organizations still lack the foundational ingredients necessary for AI success: clean data, operational discipline, governance, and clear business alignment. Every single project that OIC Advisors has undertaken for clients has resulted in some aspectof process improvement and operational maturity. Without focus in these areas, AI projects are doomed to fail. Recent reporting6 supports that enterprise AI struggles are often really data management failures disguised as AI initiatives. Without trustworthy data and repeatable workflows, AI systems become expensive experiments rather than scalable business assets.
None of this means AI lacks transformative potential - quite the opposite. AI will almost certainly reshape industries over the next decade. But transformative technologies still obey economic reality. Eventually, companies must demonstrate measurable productivity gains, revenue growth, or margin improvement. And letting AI agents loose on bad data and inefficient processes will rapidly transform small cracks into broken businesses. The better approach is to invest in the foundation – update inefficient processes, start to pay off tech debt, and align roles and responsibilities, while ensuring that your underlying data sets are clean and updated. Taking the time to do it right will save time AND money in the mid- to long-term, and allow AI pilots to quickly transition into production AI deployments.
Otherwise, at some point, boards and shareholders will stop asking, “What is our AI strategy?” and start asking a far more difficult question:
“What is the actual return?”
When that shift happens, many current AI initiatives may be exposed for what they really are — not strategic transformations, but costly exercises in corporate FOMO.
References: