Imagine investing millions in a technology that promises to revolutionize your business, only to find that 90% of similar initiatives deliver no tangible value. That’s the stark reality facing companies worldwide as artificial intelligence slips into what Gartner calls the “Trough of Disillusionment.” With global AI spending forecast to reach $2.52 trillion in 2026 – a 44% year-over-year increase – boards are now demanding answers about where all that money is going.
The Disillusionment Paradox
According to Gartner’s latest analysis, generative AI has entered the phase where initial hype fades and business leaders question return on investment. John-David Lovelock, Gartner’s chief forecaster, suggests this isn’t necessarily bad news. “They probably should be looking for AI to slip into the ditch,” he told ZDNET. “The trough is all about expectations being at their lowest. And the problems we have seen with AI in the last two years are connected to these over-the-top moonshot projects.”
MIT research supports this sobering assessment, indicating that 95% of generative AI projects fail to deliver value. But rather than signaling AI’s demise, this disillusionment phase represents a crucial maturation period. Companies are moving from speculative exploration to strategic implementation.
The Capacity Building Race
While some companies struggle with implementation, others are making massive infrastructure bets. Amazon is launching a $200 billion capital expenditure program this year, with roughly 75% allocated to AWS. This spending – which exceeds that of Google and Microsoft – focuses on expanding data centers, developing chips (Graviton and Trainium), and building AI models (Nova).
Across the globe, India’s Adani Group announced a $100 billion investment over the next decade to build AI-specialized data centers, aiming to create a $250 billion AI infrastructure ecosystem. These renewable-energy-powered facilities will support AI workloads with plans to deploy up to 5 gigawatts of capacity.
“You need to ask, ‘How deeply do I need to own this technology? How much can I deal with it as a commodity? And how much of our approach is about differentiating AI that we must own, operate, and create?'” Lovelock advises.
The Partnership Imperative
For most organizations, the path forward involves strategic partnerships rather than solo development. “This year, most people should be looking for the technology coming from their established partner stack,” Lovelock suggests. “It’s only the leaders, the visionaries, who should be looking to self-develop AI solutions or push the envelope.”
This partnership approach reflects a broader industry trend. Amazon has invested $8 billion in Anthropic and signed a $38 billion cloud deal with OpenAI, though this is dwarfed by Microsoft’s $250 billion contract with OpenAI. The key, according to Lovelock, is creating relationships where “their reward is tied to your outcome.”
Productivity Gains Emerge
Despite implementation challenges, evidence suggests AI is finally delivering measurable productivity gains. Recent analysis from Stanford University’s Digital Economy Lab indicates a U.S. productivity increase of roughly 2.7% for 2025 – nearly double the past decade’s annual average of 1.4%.
This aligns with the “productivity J-curve” theory for general-purpose technologies, where initial investment precedes measurable gains. Micro-level evidence shows AI-exposed sectors are reducing entry-level hiring by roughly 16% while augmenting skilled workers, suggesting a structural shift in how work gets done.
The Security Challenge
As AI adoption accelerates, security vulnerabilities present another layer of complexity. Recent warnings from the U.S. Cybersecurity and Infrastructure Security Agency (CISA) highlight ongoing attacks exploiting vulnerabilities in Chrome, Zimbra, ThreatSonar, and even 18-year-old ActiveX controls.
Microsoft has also warned about critical privilege escalation vulnerabilities in Windows Admin Center, rated as high risk with CVSS scores of 8.8. These security challenges underscore why many organizations prefer working with established technology partners who can provide integrated security solutions.
Market Realignment
The AI revolution is causing significant market volatility, particularly in the software sector. Investors fear that AI agents – which can perform tasks for non-technical workers – threaten traditional software-as-a-service companies by potentially becoming a new layer on top of existing software.
Yet investment continues at a staggering pace. In just the first two months of 2026, nearly 20 U.S.-based AI startups raised funding rounds of $100 million or more, continuing the trend of massive AI investments seen in 2025.
The Path Forward
So what separates successful AI implementations from the 90% that fail? Lovelock emphasizes three critical areas: “Partners, data, and processes.” He adds that success requires involving line-of-business functions and focusing on defined business outcomes.
The companies thriving in this environment aren’t necessarily those with the biggest budgets, but those making the smartest choices about capacity building, partnership strategies, and implementation focus. As the industry moves through its disillusionment phase, the winners will be those who view this period not as a setback, but as an opportunity to build sustainable AI strategies that deliver real business value.