The Big Short Michael Burry: Company A's prospectus reveals "catastrophically fragile" systemic risks in the AI boom

The "Big Short" figure Michael Burry stated via his Substack account that Anthropic’s newly public IPO S-1 filing reveals catastrophic systemic risks in the broader AI boom, not limited to Anthropic alone. His core arguments are as follows: 1. Massive compute commitments: Anthropic has signed approximately $537 billion in compute and infrastructure purchase/lease agreements, primarily with Amazon, Google/Alphabet, Microsoft, Broadcom, and other firms. Pending deals (including those with SpaceX AI) could add hundreds of billions more, bringing the total close to $800 billion. This is roughly 12 times its current annual revenue run rate of around $65 billion, and over 20 times its trailing 12-month revenue. 2. Circular and interconnected financing structure: Major tech giants act as Anthropic’s lenders, suppliers, distributors, and shareholders simultaneously. The financing arrangement is described as "garish and audacious". This high interconnectivity appears as a safety net during an upcycle, but once growth stalls or enters a downturn, the entire ecosystem becomes extremely fragile. He draws a parallel to vendor financing during the dot-com bubble. 3. Slowing growth + accelerating costs: Quarterly revenue growth dropped from ~144% in the prior quarter to roughly 46%, while compute costs surged by ~78%. Burry notes the S-1 data is as of June 30, failing to fully reflect changes in the developer ecosystem in Q3 and beyond, adding "nothing changes my overall view". He argues Anthropic’s structure works when AI grows at a sustained rapid pace, but any significant slowdown could trigger a chain reaction or even collapse. The risk, he emphasizes, lies in the system’s "circularity and interconnectedness", not just technical issues.