The U.S. government has partially eased restrictions on the use of Anthropic's top-tier artificial intelligence model, Mythos 5, limiting access exclusively to domestic companies.
This regulatory adjustment comes approximately two weeks after the Donald Trump administration announced export controls on Anthropic's AI models.
However, South Korean companies, including Samsung Electronics and SK Hynix, are expected to remain restricted from using Mythos 5.
According to Reuters and The Washington Post on June 26 (local time), the U.S. Department of Commerce notified Anthropic in a private letter that only specific companies within the United States are permitted to use the AI model Claude Mythos 5.
In the letter obtained by the media, U.S. Secretary of Commerce Howard Lutnick stated, "Sufficient security measures have been put in place for certain trusted partners to use the Mythos 5 model."
In effect, the U.S. government has mandated that only specific domestic companies and institutions it has verified as safe may utilize Mythos 5.
The letter also noted that the U.S. government reserves the right to modify the list of approved entities at any time.
Specific details regarding the approved entities have not yet been confirmed.
However, The Washington Post, citing anonymous sources, reported that approximately 100 companies and institutions within the U.S. are expected to be included.
In a statement released following the media reports, Anthropic said, "Today, the government notified us that we may resume deployment of Mythos 5, our most powerful cybersecurity model, to various U.S. agencies that operate and defend critical national infrastructure."
This is interpreted to mean that Mythos 5 will be used only by verified U.S. companies and institutions, while the restrictions on foreign companies remain in place.
Consequently, South Korean firms such as Samsung Electronics, SK Hynix, and SK Telecom, which were previously known to have early access to Mythos—the predecessor to Mythos 5—through collaboration with Anthropic, are unlikely to regain access.
Mythos 5 is a next-generation AI model specialized in cybersecurity released by Anthropic.
The model possesses powerful capabilities in detecting and defending against security vulnerabilities.
However, it has also been pointed out that the model is a double-edged sword, as it is difficult to counter with existing cybersecurity systems.
For this reason, Anthropic launched Project Glasswing, a global AI security cooperation project, and provided the pre-release Mythos model to verified companies and institutions in various countries.
The intention was to proactively identify software vulnerabilities through collaboration with these partners.
South Korean companies, including Samsung Electronics, were also known to have been granted access to Mythos upon joining Project Glasswing on June 2.
However, after Anthropic officially launched Mythos 5 and Fable 5 based on Mythos, the U.S. government announced export control measures targeting these cutting-edge AI products, effectively blocking access for South Korean companies.
The U.S. government's control measures targeting advanced AI services are becoming evident with other companies as well.
OpenAI unveiled its next-generation AI model, GPT-5.6, on this day.
However, instead of making it available to everyone as in the past, the company decided to provide it exclusively to select institutions designated by the U.S. government.
This has drawn criticism that the U.S. government is arbitrarily controlling access to cutting-edge AI models.
Within the U.S. AI industry, some point out that the government's unusual and broad export controls on advanced AI models could actually undermine U.S. leadership in AI.
Meanwhile, The New York Times reported, citing anonymous sources, that Anthropic and the U.S. government are also discussing the issue of restoring access to Fable 5, which has drawn significant interest from consumers.
Unlike Mythos 5, which is permitted only for computing infrastructure companies and a small number of institutional clients, Fable 5 was released for general users by applying guardrails (safety mechanisms).
In other words, the Mythos model was provided to professional companies and institutions as a product named Mythos 5, and to individuals as a product named Fable 5.
Previously, on June 12, the Trump administration issued export control guidelines to completely block all foreign nationals from accessing Mythos 5 and Fable 5, citing national security concerns.
As a result, Anthropic, which lacked the functionality to verify the nationality of its users, had no choice but to block all users from accessing the two models.
U.S. media reported that employees working at the U.S. companies and institutions granted access this time are now able to use these models, even if they are foreign nationals.
※
Copying, redistribution, and unauthorized use in AI training are strictly prohibited.