Something unusual happened at the World Economic Forum in January 2026. Larry Fink, the CEO of BlackRock, the world's largest asset manager, stood in front of the global elite at Davos and essentially told them that capitalism was failing. AI, he warned, could be its next great failure. It was a striking moment. Not because the critique was new, but because of who was delivering it. When the person managing $11 trillion in assets starts questioning the system that made that possible, it is worth paying attention.

The wealth engine that only works for some

Fink's argument was direct. Since the fall of the Berlin Wall, more wealth has been created than at any prior point in human history. But in advanced economies, that wealth has gone to a shrinking slice of the population. AI, he said, is on track to repeat and accelerate that pattern. "Early gains are flowing to the owners of models, owners of data, and owners of infrastructure," Fink told the Davos audience. "The open question: What happens to everyone else if AI does to white-collar workers what globalization did to blue-collar workers?" The numbers back him up. In 2025, a group of 34 AI-related stocks surged over 50%. The median increase in net worth among the 50 wealthiest Americans was nearly $10 billion. Google co-founders Larry Page and Sergey Brin got $101 billion and $92 billion richer, respectively. Meanwhile, the bottom half of Americans own roughly 1% of stock market wealth. This is the K-shaped economy in action. The top accelerates upward. Everyone else flatlines or falls behind.

The labor problem

The IMF estimates that nearly 40% of global employment is exposed to AI. In advanced economies, the figure rises to about 60%. What makes AI different from previous waves of automation is its reach into high-skilled, white-collar work, the kind of work that was supposed to be safe. Roughly half of those exposed jobs may benefit from AI integration through enhanced productivity. But the other half could see reduced hiring, lower wages, or outright elimination. Nobel laureate Geoffrey Hinton put it bluntly: "What's actually going to happen is rich people are going to use AI to replace workers. It's going to create massive unemployment and a huge rise in profits." This is not speculative anymore. Economist Jason Furman estimated that U.S. GDP growth in the first half of 2025 was almost entirely driven by AI-related investment, primarily data centers and associated infrastructure. Strip that out, and growth was essentially flat at 0.1%. OECD researchers went further, suggesting the U.S. would have been in outright recession without AI capital spending. So the economy is growing, but the growth is coming from capital investment, not from broadly shared prosperity. The spending is building infrastructure that, by design, reduces the need for human labor.

The capitalism paradox

This is where the deeper tension emerges. Capitalism depends on a cycle: businesses pay workers, workers spend money, spending generates demand, demand sustains businesses. AI threatens to break that cycle by making it possible to produce more with fewer workers. Nobel laureate Joseph Stiglitz frames AI as a textbook case of how technology can turbocharge inequality. "If we don't do anything about managing AI, there is a threat that it will lead to more inequality," he told Fortune in March 2026. "And since inequality is such a bad, serious problem in our society, that is a great concern to me." What concerns Stiglitz most is not AI itself, but the political context surrounding it. The same people driving AI adoption are simultaneously pushing to shrink the government institutions that could manage the transition. "They are creating the conditions that make it impossible for a successful AI transition," he argued. Stiglitz draws a parallel to the Great Depression, which was partly driven by a success story gone wrong: agricultural productivity surged, millions of farm workers were no longer needed, and the economy had no mechanism to absorb them. It took World War II and massive government intervention to resolve that crisis. Today, we face a similar structural challenge with no comparable institutional response in sight.

Power concentration and "AI capitalism"

Beyond labor displacement, AI is driving an unprecedented concentration of economic power. Researchers have described this as "AI capitalism," characterized by three reinforcing dynamics: the commodification of data, the extraction of value from that data by a handful of firms, and the concentration of AI talent and compute capacity among those same firms. This follows the "winner takes all" logic. Building competitive AI systems requires enormous datasets, vast computing infrastructure, and highly specialized researchers. The barriers to entry are staggering. The result is a market structure where a few companies, primarily the large American tech platforms, control the infrastructure of an increasingly AI-dependent economy. The academic Jathan Sadowski and others have argued that this dynamic turns AI into an "ultimate weapon for capital," one that not only automates production but centralizes the means of production in ways that earlier technologies never could.

The case for a different path

None of this is inevitable. The distinction Stiglitz draws between AI and what he calls "IA," intelligence assisting, is useful here. AI as a displacement engine concentrates gains at the top. IA as a tool that augments human capabilities can distribute benefits more broadly. Stiglitz uses AI in his own research. "I view AI as augmenting my abilities," he said. "It's sort of like having a team of research assistants, but faster." He compares it to the microscope or telescope: tools that extended human perception rather than replacing human judgment. The IMF has proposed an AI Preparedness Index, measuring countries' readiness across digital infrastructure, labor-market policies, innovation capacity, and regulation. Their findings show that wealthier nations are better positioned, but there is significant variation. The countries performing best, like Singapore, the United States, and Denmark, tend to combine technological investment with strong social safety nets and adaptive regulatory frameworks. Some concrete policy directions are emerging:

Where this leaves us

The honest answer is that we do not know how this plays out. AI is a general-purpose technology, the kind that reshapes entire economies over decades. The steam engine, electricity, and the internet all followed this pattern, each creating enormous wealth while also causing significant disruption and displacement before new equilibria were found. What is different this time is the speed and the scope. AI is advancing faster than previous general-purpose technologies, and it reaches into cognitive work that was previously thought to be uniquely human. The window for proactive policy is narrower than it was during the Industrial Revolution or the rise of the internet. Fink closed his Davos remarks with a challenge: "Many of the people most affected by what we talk about here will never come to this conference. That's a central tension of this forum. Davos is an elite gathering trying to shape a world that belongs to everyone." That tension, between who builds the future and who lives in it, is the defining question of AI and capitalism. The technology itself is neither liberating nor oppressive. The systems we build around it will determine which one it becomes.

References