By: Sharon Goldman
Originally published on Fortune, October 1, 2025.
Over the nearly three years since ChatGPT’s launch in November 2022, generative AI has created a frenzy that has radiated like the midday summer sun—hot and unrelenting.
And for the AI companies rocketing forth like heat-seeking missiles, including OpenAI, Anthropic, Google, Microsoft, Meta, and xAI, the sun is still shining: The research firm Gartner forecasts worldwide AI spending will reach nearly $1.5 trillion in 2025 and surpass $2 trillion in 2026, fueled by integration into smartphones, PCs, and enterprise infrastructure. Elon Musk and other AI leaders continue to insist that artificial general intelligence (AGI)—an AI that can think and learn like a human, across many tasks—is on the horizon.
But on the ground, the temperature is dropping, and it’s starting to feel like sweater weather. Among customers and in financial markets, skepticism is rising as some question whether the massive investment in AI will ever be justified by revenues. Startup funding is under sharper scrutiny for small and midsize firms; enterprise projects are stuck in “pilot purgatory”; corporate buyers are questioning return on investment for AI expenditures; and the rising cost of computing power has become a wall many would-be competitors can’t climb.
We don’t yet know whether this chill will eventually turn into an “AI winter,” the industry term for the stage in past AI hype cycles when enthusiasm waned and investment dried up. As my colleague Jeremy Kahn has noted, AI winters have often followed a familiar arc: promising advances that failed to deliver, leaving those footing the bill disillusioned. Sometimes the trigger was academic research exposing the limits of certain techniques. Sometimes it was the failure of real-world adoption. Most often, it was both.
“There are certainly a few autumnal signs, a falling leaf carried on the breeze here and there, if past AI winters are any guide,” Kahn recently wrote. Only time will tell whether this is “the prelude to another arctic bomb that will freeze AI investment for a generation, or merely a momentary cold snap before the sun appears again.”
To continue reading this article, click here.