Anthropic says it was directed by the US government to suspend access to Claude Fable 5 and Claude Mythos 5 for foreign nationals, including foreign-national employees inside the company. The directive, which Anthropic says arrived at 5:21 p.m. ET on June 12, forced the company to disable both models globally to ensure compliance.
The company’s official statement frames the order as an emergency export control action tied to national security concerns. Other Anthropic models, including Claude Opus 4.8, are not affected and remain operational.
The immediate market reaction was visible in pre-IPO-linked trading. The Anthropic perpetual contract on Hyperliquid reportedly fell 3.7% to about $1,627, down from post-launch highs above $1,800, with open interest around $8.6 million.
Anthropic Pushes Back On The Government’s Evidence
Anthropic said the directive followed reports of a non-universal jailbreak vulnerability in Fable 5. The company argued that the technique described to it did not justify a full recall-style shutdown of a commercial model.
In its statement, Anthropic said the government had provided only verbal evidence of a narrow jailbreak, describing it as a prompt that asked the model to review a specific codebase and identify software flaws. Anthropic said those flaws were minor, previously known, and could also be found by other public models without requiring a bypass.
That distinction matters because the standard being applied here could affect more than Anthropic. The company warned that if the same threshold were applied across the industry, it would essentially halt all new model deployments for frontier AI providers.
Why Crypto Markets Are Watching AI Pre-IPO Tokens
The story matters for crypto because pre-IPO markets and perpetual-style exposure have turned private tech names into tradable sentiment instruments. Anthropic is not a crypto company, but its market-linked contracts allow crypto-native traders to react quickly to AI-sector regulatory news.
The risk is that these markets can move sharply on incomplete information. The government’s technical report has not been made public, and Anthropic says it has only received verbal evidence so far. That leaves investors weighing a company statement against an opaque national-security process.
The key point is that AI regulation is becoming a tradable event. When frontier-model access can be restricted by government directive, private-market valuations and tokenized exposure products can react almost immediately.
The wider signal is that AI infrastructure is becoming part of the same speculative market map as crypto, private equity and tokenized exposure. When a model access decision changes perceived company value, traders can now express that view almost instantly through pre-IPO-linked instruments.
The risk is that these instruments can react faster than the public evidence base develops. Until the government releases more detail, the market is pricing uncertainty around Anthropic’s product access, regulatory exposure and frontier-model deployment risk.
