The Trump administration has mandated that Anthropic PBC must obtain US government approval before allowing foreign nationals to access its Fable 5 and Mythos 5 artificial intelligence models. This directive, issued via a Commerce Department order by Secretary Howard Lutnick, marks an unprecedented expansion of export control laws to cover the usage of cloud-based AI systems. The move, which cited national security concerns regarding potential 'jailbreaks' or unauthorized access to guardrails, led Anthropic to immediately disable access to the specified models for foreign users.

Previously, the technology sector operated under the assumption that export controls applied to the transfer of software code but not to the usage of cloud-based programs. Commerce Department officials and legal analysts noted that this order treats the simple usage of an AI model as a technology transfer, a classification that remains untested in courts and runs counter to established regulatory practices. The Bureau of Industry and Security is now positioned to apply similar oversight to other AI companies, raising significant concerns among developers and customers regarding potential service disruptions.

In response to the mandate, Anthropic and administration officials are holding ongoing discussions to address the identified national security risks. At the recent Group of Seven summit in France, international leaders discussed the implications of this intervention, with calls for greater technological sovereignty increasing among global partners. Industry executives, including Cohere Inc. CEO Aidan Gomez, have identified this event as a catalyst for firms to diversify their AI supply chains to reduce reliance on centralized providers. The administration's intervention represents a significant shift from the voluntary participation frameworks outlined in an executive order signed by President Donald Trump approximately two weeks prior to the directive.

The Trump administration has mandated that Anthropic PBC must obtain US government approval before allowing foreign nationals to access its Fable 5 and Mythos 5 artificial intelligence models. This directive, issued via a Commerce Department order by Secretary Howard Lutnick, marks an unprecedented expansion of export control laws to cover the usage of cloud-based AI systems. The move, which cited national security concerns regarding potential 'jailbreaks' or unauthorized access to guardrails, led Anthropic to immediately disable access to the specified models for foreign users.

Previously, the technology sector operated under the assumption that export controls applied to the transfer of software code but not to the usage of cloud-based programs. Commerce Department officials and legal analysts noted that this order treats the simple usage of an AI model as a technology transfer, a classification that remains untested in courts and runs counter to established regulatory practices. The Bureau of Industry and Security is now positioned to apply similar oversight to other AI companies, raising significant concerns among developers and customers regarding potential service disruptions.

In response to the mandate, Anthropic and administration officials are holding ongoing discussions to address the identified national security risks. At the recent Group of Seven summit in France, international leaders discussed the implications of this intervention, with calls for greater technological sovereignty increasing among global partners. Industry executives, including Cohere Inc. CEO Aidan Gomez, have identified this event as a catalyst for firms to diversify their AI supply chains to reduce reliance on centralized providers. The administration's intervention represents a significant shift from the voluntary participation frameworks outlined in an executive order signed by President Donald Trump approximately two weeks prior to the directive.