The Overindulgence of AI Spending: A War-Level Burn Without Clear Returns

It’s the season of excess, and the tech industry has truly outdone itself. This year, approximately $1.5 trillion has been poured into AI development—an amount comparable to the cost of major U.S. wars in Iraq and Afghanistan from 2001 to 2014, which totaled around $1.5 to $1.7 trillion in direct spending. According to Gartner, global AI investment is projected to hit nearly $1.5 trillion this year, placing this AI boom right alongside wartime spending in magnitude.

The Missing Piece: Returns on Investment?

What’s conspicuously absent from this spending frenzy is a clear justification. As with wartime expenditures, some are starting to question the rationale behind such massive investments into a technology that remains perishable and uncertain in terms of long-term gains. A recurring question emerges: How many years will it take before we see a return on this AI investment?

It’s reminiscent of closing time at a buffet—despite the abundance, the hope that the feast will turn profitably is waning. Now, corporate executives are inundating press releases with buzzwords like “AI” and “agents” to keep the momentum alive—even if their understanding of the technology remains superficial.

Corporate Speak and the AI Hype Machine

Larry Feinsmith, head of Global Tech Strategy at JPMorgan Chase, recently stated:

“In the era of AI and agents, the benefits and value will be enormous, but so is the complexity.”

His words—coming from a Wharton-educated executive—serve as a cautionary note amid the hype. He's advising companies to buy into AI, promising vast future gains, yet offering little clarity on when those benefits will materialize.

The Economic Reality

Economist Torsten Slok from Apollo Global Management highlighted the precariousness of current corporate spending. He noted that outside of AI, there’s been virtually no growth in capital expenditure, and some believe AI investments are the key factor keeping the U.S. economy afloat amid fears of recession.

Despite widespread enthusiasm, the actual use cases for AI remain limited mostly to chatbots and speculative “agents,” with predictions of a revolution comparable to electricity but little tangible application beyond automation and customer service.

Infrastructure Booms and Potential Busts

Tech giants like Microsoft are investing in sprawling AI data centers—massive, continent-spanning hyperscale infrastructures. While these “datacenter superclusters” symbolize ambition, Goldman Sachs warns that this AI infrastructure surge could end poorly if the ROI fails to justify the enormous investments.

Executives at companies like Dell and Nvidia are bullish, predicting an AI revolution, yet actual use cases are scarce, and the promised transformations remain elusive. A few years from now, the predicted AI “agents” may still be mostly hype.

The AI Agent Dilemma

Salesforce CEO Marc Benioff claims his AI agents are already assisting customers—but studies, such as one from Carnegie Mellon University, suggest that AI agents fail about 70% of the time. The picture resembles efforts to put family favorites or relics into important roles—they might succeed or fail, but ultimately, everyone loses if they don’t deliver.

The Business Case for AI: Only Partial Success

According to Forrester Research, AI must demonstrate real value to keep companies investing. Fearing stagnant or declining returns, 25% of enterprises are delaying AI projects into 2027. Companies are beginning to realize that despite the billions spent, the impact on essential financial metrics—like EBITDA—remains minimal.

Final Thoughts

You might spell “AI” with an 'A' for ambition, but in many cases, it fundamentally affects only the “IT” department—marking a shift from meaningful profit-driving to “ITDA,” or “IT-driven activities,” rather than real earnings growth. Until strong, demonstrable returns emerge, the industry’s spending binge may continue to resemble an overextended buffet—plentiful, yet increasingly hollow.

This article underscores the disparity between the massive investments in AI and the tangible, proven business results, questioning whether the current AI spending spree is sustainable.