Steve Eisman warns AI boom faces structural Achilles heel as OpenAI and Anthropic concentration risk threatens hyperscaler cloud revenues

Steve Eisman, the investor known for his prescient bet against the US housing market ahead of the 2008 financial crisis, has warned that the artificial intelligence boom contains a structural vulnerability: approximately 70% of AI-related revenue at Microsoft, Amazon, Alphabet (Google's parent company), and Oracle is generated by just two companies, OpenAI and Anthropic. Those two AI startups also account for as much as 25% to 35% of total cloud revenue at those hyperscalers (the term for companies that operate very large-scale cloud computing infrastructure). "The futures of these massive companies, in a sense, are a bet that OpenAI and Anthropic are going to succeed," Eisman said on CNBC's Fast Money in mid-August 2026. He identified Chinese open-source AI models as the primary competitive threat: cheaper, open-weight (publicly accessible) models from China are gaining market share, and a price war could erode the value of the proprietary model ecosystem on which hyperscaler AI revenues depend. Michael Burry, another investor associated with the Big Short trade, has separately taken an even more bearish view, questioning whether current and future AI demand is genuinely end-customer-driven or whether a significant portion is circularly financed. Burry has disclosed bearish positions against and the broader semiconductor sector. For lawyers and law firms, the Eisman and Burry warnings are commercially material: firms that have made significant commitments to AI infrastructure deals, AI-linked financing structures, or technology licensing arrangements with hyperscalers as counterparties need to assess the concentration risk embedded in those engagements. The debate also affects the legal profession's own AI investment decisions, given that law firms are increasingly spending on AI tools powered by models from OpenAI and Anthropic.