Anthropic abruptly suspended access to its Fable 5 and Mythos 5 AI models following a US government export control directive that bans foreign nationals from using them. The company confirmed the move was compelled by authorities, according to a statement to The Record.
The directive, which invokes national security powers, is believed to be the first instance of such authorities being applied to restrict AI software rather than hardware like semiconductors or chips. The scope of the ban and affected user base remain unclear, but the action signals an escalation in export controls beyond traditional physical goods.
Dark Reading reports that Anthropic received the directive and suspended access immediately. The company has not disclosed whether the models were designed for cybersecurity applications, though The Record notes the models were described as cybersecurity-focused.
No specific technical details about the models—such as capabilities, training data, or potential abuse vectors—have been released. Neither Anthropic nor the US government has provided a timeline for resolving the restriction or criteria for lifting the ban.
The move highlights growing tension between AI development and national security, as regulators grapple with dual-use technologies. Critics argue such broad directives could stifle innovation without clear evidence of harm, though the government has not publicly detailed specific threats.