A year ago, CEO skepticism about AI would have tanked a stock price. Now, analysts and investors are questioning whether a more measured approach makes sense.

Third-quarter earnings calls revealed a dramatic shift in sentiment. While AI mentions increased just 4.8% (from 292 to 306 calls among S&P 500 companies), questions about an AI bubble skyrocketed 740% between Q2 and Q3. Translation: investors are growing as concerned about losing money on AI as they are about making it.

This shift reflects a sobering reality check. The AI frenzy has abated significantly, and CEOs are now on safer ground, being cautious rather than being perceived as missing the wave.

The statistics paint a stark picture:

• Most AI projects fail to scale past the pilot stage
• Of those that do scale, only 20% generate meaningful ROI
• More than 40% of AI agent projects are projected to fail by 2027
• Few companies outside the tech giants can point to concrete use cases generating real business impact

Leaders who invested heavily in AI are realizing returns will take considerably longer than initially predicted. They’re rethinking investment strategies and adjusting forecasts accordingly.

Being vocal about a possible AI bubble serves two purposes: managing investor expectations and demonstrating careful balance between investment and profitability. With markets strong and most companies performing well, leaders feel comfortable backing away from what was a “frothy period” for AI investment.

Importantly, an AI bubble differs from an overall financial bubble. CEOs and CFOs can discuss AI concerns without appearing to ignore the next great tech advance.

For investors, the math doesn’t add up: spending billions to save employees a few hours per week generates questionable value, especially when it’s unclear how employees use that extra time.

Generating real returns requires the granular work of changing roles, teams, and workflows to incorporate AI. Most companies aren’t doing that yet. Until they do, the gap between AI investment and AI returns will continue to widen, and investor skepticism will continue to grow.

The message is clear: the easy part was buying the technology. The hard part, organizational transformation, is where most companies are still stuck.