• The U.S. government has ordered Anthropic to suspend foreign access to its AI models Fable 5 and Mythos 5.
• This measure is in response to national security concerns, reportedly linked to potential "jailbreaking" methods.
• Anthropic views the government's action as a misunderstanding and is seeking to restore access soon.
• The order escalates U.S. export control efforts regarding AI technology.
• Anthropic has previously opposed its technology being used for surveillance and military purposes.

The U.S. government has ordered Anthropic to suspend foreign access to its most advanced AI models, Fable 5 and Mythos 5, citing national security concerns related to potential methods for "jailbreaking" the systems. On June 12, Anthropic received a directive from U.S. Commerce Secretary Howard Lutnick, mandating the abrupt disabling of these models for all foreign nationals, including foreign employees at the company, according to Channel News Asia and Reuters.

In its response, Anthropic stated, "The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance." They also highlighted that they understand the government's concerns stem from a belief that a "jailbreaking" method could allow Fable 5 to be exploited for identifying software vulnerabilities. Anthropic contended that the order fails to consider their assessments, which indicated no universal hack method had been confirmed, according to Le Monde and Channel News Asia.

This order marks a significant escalation in the U.S. efforts to control the export of AI technologies deemed risky. Previously, U.S. export controls have focused primarily on hardware and tools rather than directly limiting foreign access to AI models themselves. Anthropic's situation is further complicated by past disputes with government officials regarding the potential military application of its technologies, leading to the company being placed on a supply chain blacklist, as noted in reports from Le Monde and Channel News Asia.

Anthropic's recent launch of the Fable 5 model had been intended to advance its AI capabilities; however, the national security intervention has now required the rapid disabling of access to this model. The company believes that the regulatory actions do not reflect principles of fair regulation and could stall the deployment of new models across the industry, as stated in their communications with the U.S. government reported by Reuters and Le Monde.