From Open Internet to Open Intelligence: Why AI’s Market Structure Matters More Than Ever

In this piece for Brookings’ Tech Tank, author Tom Wheeler discusses how the global AI market is rapidly consolidating around a small group of dominant firms that already control data, computing power, and digital distribution. He argues that AI is not just another technology cycle but a foundational infrastructure transformation—comparable to electrification or the internet—emerging at a time when market power is already highly concentrated.

Wheeler warns that the same companies that once championed openness during the early internet era are now repeating a familiar pattern: using openness to grow, then closing off systems to entrench dominance. As these firms transition from Big Tech to “Big AI,” they are importing strategies that restrict competition, raise switching costs, and externalize social and economic risks onto the public.

The piece emphasizes that the true value of AI lies not in ever-larger models, but in the diffusion of applications across the economy. Productivity gains, innovation, and public benefits depend on broad access to AI capabilities. That diffusion, Wheeler argues, will fail if a handful of companies control critical bottlenecks, making the prevention of market chokepoints a central AI policy objective.

Wheeler identifies three decisive chokepoints—data, compute, and models—where concentration threatens openness and competition. Restricted access to data, dominance over cloud computing infrastructure, and closed foundation models controlled through proprietary APIs all risk slowing innovation and concentrating power. Without policies that promote interoperability and prevent gatekeeping, Wheeler posits, the AI era will be defined by control rather than competition.

CONTINUE READING…