We Haven’t Entered an ‘AI Winter,’ But We Need Strategies to Weather the Chill

Since the launch of ChatGPT in November 2022, nearly three years ago, the generative AI wave has been blazing hot and persistent, like the midday sun in midsummer. For rapidly evolving AI companies such as OpenAI, Anthropic, Google, Microsoft, Meta, and xAI—practically acting as heat-seeking missiles—the momentum shows no signs of slowing down. Research firm Gartner predicts that as AI technologies integrate into smartphones, personal computers, and enterprise infrastructure, global AI spending will approach $1.5 trillion by 2025 and surpass $2 trillion by 2026. Elon Musk and other AI leaders continue to insist that Artificial General Intelligence (AGI)—AI capable of human-like reasoning and learning across multiple tasks—is within reach.Yet in practical scenarios, the “heat” is quietly fading, as if we’ve entered a season where a sweater is needed to stave off the chill. Among clients and financial markets, skepticism is rising. Questions are being asked: will massive AI investments ultimately yield reasonable returns? Funding for small and mid-sized startups is facing more stringent scrutiny; enterprise AI projects are trapped in “pilot purgatory”; corporate buyers are beginning to question AI investment ROI; and rising computing costs have become a barrier too high for many potential competitors.It remains unclear whether this chill will evolve into a full-blown “AI winter”—a term that refers to the periods in past AI hype cycles when enthusiasm waned and investment dried up.

As my colleague Jeremy Kahn noted, AI winters often follow a familiar pattern: seemingly promising technological advancements fail to deliver, disappointing investors. Sometimes the trigger is academic research revealing limitations of a specific technology; other times, it is practical applications that falter. Often, both occur simultaneously. “If we look to previous AI winters, there are indeed a few signs of autumn today—sporadic leaves drifting in the wind,” Kahn recently wrote. Whether this is the prelude to “another severe storm that will freeze AI investments for a generation” or merely “a brief cold snap before the sun returns,” only time will tell. The latter scenario may not be entirely negative.Ron Colleon, chief analyst at Forrester Research, told Fortune that he views the industry as undergoing a “necessary adjustment.” “Our thermometer was previously malfunctioning,” he said. “Now we can finally measure the temperature accurately.” Colleon emphasized that enterprise clients are not abandoning AI. On the contrary, faced with overhyped promises, they are recalibrating their strategies. For instance, marketing around agentic AI once implied that every company needed to deploy general AI agents for all employees overnight. “Now companies are saying, ‘We don’t necessarily need AGI for everyone tomorrow,’” he explained. “We need to think more carefully about data architecture and content quality to take a more holistic approach.”Clearly, the feverish expectation of fully achieving AGI by 2027 has cooled.

But this does not imply that AI investment is diminishing. Colleon sees a more pressing issue: “a gap between leadership expectations and actual outcomes.” Executives often issue directives disconnected from real business objectives, such as “every employee must use generative AI twice daily.” “Disappointment quietly builds,” he explains—not because AI has failed outright, but because these expectations were misaligned with practical applications from the start.Deloitte CTO Bill Briggs also acknowledges a shift in AI sentiment but says the current situation is far from the peril seen during the late-1990s tech bubble burst. “We are at an inflection point, but I don’t believe we will repeat the Internet bubble crash,” he said, noting that AI continues to drive transformation and new business models are emerging. Overall, Briggs observes that AI is transitioning from a rising star to a behind-the-scenes operational powerhouse, quietly reshaping how companies approach processes, products, and decision-making. “AI’s trajectory may be akin to electricity—ubiquitous yet invisible, powering everything around us.”However, not everyone agrees that “the industry’s heat is waning.” Steve Hall, EMEA partner, president, and chief AI officer at global technology research and consulting firm ISG, insists that the likelihood of an AI winter is minimal. “It’s early spring,” he says. “Generative AI has been around less than three years, and agentic AI only 15 months.

While the hype cycle has peaked, many sectors are just beginning to show budding signs.” Hall notes that current investments focus on chips and hyperscale enterprises. Over the past three years, these large tech and cloud companies have been building infrastructure to support their AI projects. Meanwhile, SaaS vendors are investing in “agentizing” their applications in 2024 to drive intelligent upgrades to business processes.Regarding skeptics’ claims of “technology application stagnation,” Hall sees this as a natural experimental phase. “We view these pilot projects not as failures at scale but as essential testing and validation steps before committing valuable resources. For such an exciting technology, this is exactly how enterprises should respond,” he says.Overall, the AI chill may recede or intensify. Regardless, history shows that hype alone cannot sustain momentum; progress must be incremental. For executives seeking clarity amid the noise, the key is not the current stage but how to wisely steer AI investments. Experts outline four strategies to navigate the chill:Anchor AI to Strategy: Ron Colleon warns against chasing short-term wins—like shaving a few seconds off call-center response times or sending bulk sales emails—which rarely create lasting value. “

If these initiatives are not linked to actual efficiency gains, effectiveness improvements, or transformational goals, they will fail,” he says. True success comes to companies that directly tie AI pilots to measurable outcomes.Master Business Communication: Deloitte’s Bill Briggs emphasizes that leaders securing funding for new AI projects don’t focus solely on technology. They position AI as a driver of business growth. “CEOs need to see you as a business-savvy partner with tech expertise, not just a technologist who occasionally talks business,” he told Fortune. This means connecting AI projects to results that capture executives’ attention: expanding markets, improving customer satisfaction, optimizing operations, and building sustainable competitive advantages.Leverage Ecosystems: Steve Hall stresses that enterprises should integrate into the broader AI ecosystem rather than attempting to build all capabilities in-house. “This is not a task that can be accomplished alone,” he says, particularly given the foundations laid by hyperscalers, chipmakers, and SaaS vendors.Balance Vision with Practical Innovation: “My advice to tech leaders is to lead with curiosity and optimism, but always keep one hand firmly on the practical steering wheel,” Briggs says. “Industry landscapes shift rapidly. The goal shouldn’t just be ‘adopt AI,’ but to fully embed AI into operational frameworks.”