The AI Bubble Bursts: CIOs Learn a Hard Lesson in Cost Control

The fervor around artificial intelligence has defined the modern corporate landscape for years. Chief Information Officers (CIOs) and other technology leaders have touted AI as a panacea for efficiency gains and innovation, but a growing number of companies are now facing a harsh reality: the costs associated with widespread AI adoption are soaring without delivering corresponding value.

Samsara, a company at the forefront of this trend, has seen significant benefits from its AI investments. However, CIO Stephen Franchetti has recently implemented caps on usage for non-technical employees, reflecting a growing recognition of the need for cost control.

As global AI spending reaches $2.5 trillion this year – a 44% increase from last year’s levels – CIOs and other technology leaders are facing a delicate moment in their AI journey. After years of promoting AI adoption, some companies are now tightening up how frequently these tools can be used and retraining staff to better understand smaller and cheaper AI models that can handle many workplace tasks.

The realization that AI is not always as effective or efficient as promised has driven this change. “2026 is the year everyone finds out that AI is actually really hard,” says Will Sommer, a quantitative modeling expert at Gartner. “It’s not a free lunch. It requires a lot of thought and effort to get right.” Companies can easily spend thousands of dollars per head on AI tools whose output is essentially junk and doesn’t bolster productivity.

Gartner has issued several reports highlighting the risks associated with AI adoption, including a bearish report warning that AI coding costs would overtake the average developer’s salary by 2028 due to rising token consumption and consumption-based fees. The implications are stark: as AI investments continue to soar, companies may soon face financial reckoning.

The Limits of AI Adoption

CIOs and technology leaders must adopt a more nuanced approach to AI adoption. Rather than simply throwing money at the latest AI tools, companies must carefully consider their needs and goals before investing in AI.

Sagnik Nandy, CTO of Docusign, has taken this seriously both internally and externally. Internally, he has adjusted his company’s AI coding agents to reduce token usage by almost 50% by limiting the context they pull from. Externally, he advocates for more careful consideration of AI model usage and advises business leaders to manage their digital spending like any other line item.

Industry leaders are also recognizing the need for a more measured approach to AI adoption. Jim Dausch, Chief Digital and Technology Officer at Yum Brands, notes that “a vast majority of tasks” can be handled by basic, less expensive models. As such, his company promotes training on AI model usage and advises business leaders to closely manage their digital spending.

The Way Forward

The future of AI adoption is clear: companies must approach AI with a more critical eye. Rather than relying on flashy tools and promises of efficiency gains, they must carefully consider their needs and goals before investing in AI.

Katya Andresen, Cigna’s Chief Data, Digital, and AI Officer, notes that “the way you really run up costs is by using the most expensive models with no guardrails around them.” This cautionary tale highlights the need for companies to adopt a more multimodal approach to AI adoption – one that balances the benefits of advanced AI tools with the need for cost control.

The Next Chapter

As CIOs and technology leaders navigate this complex landscape, they would do well to heed Will Sommer’s words: “AI is not a free lunch. It requires a lot of thought and effort to get right.” By adopting a more measured approach to AI adoption – one that balances innovation with cost control – companies can avoid the pitfalls of over-investment in AI tools.

The AI bubble may have burst, but its legacy will continue to shape the corporate landscape for years to come.