Encryption, spyware, and now Mythos: History shows why cyber export control doesn’t work

The White House ordered Anthropic to restrict export of its advanced AI models Fable and Mythos amid national-security concerns, prompting the company to suspend external access; the directive followed reports of access granted to a South Korean telecom suspected of China ties and claims Amazon researchers found a way to bypass model safeguards. The article situates this action within a history of export-control attempts (e.g., PGP and the Wassenaar Arrangement), highlights limitations and enforcement challenges of controlling dual-use cyber technologies, and argues such controls are unlikely to fully prevent malicious use while raising compliance burdens for U.S. AI firms.