Anthropic is conducting direct negotiations with US regulatory authorities regarding access to its advanced AI model 'Mythos', currently blocked to Korea under US export controls. Kim Yong-beom, Blue House Policy Office Director, revealed on the 24th (local time) that the Ministry of Science and ICT continues dialogue with Anthropic, which is simultaneously talking with Washington D.C. regulators. The US government classifies Mythos as strategic technology directly linked to national security, imposing strict access restrictions despite Anthropic's stated preference to make the product available to all customers.
Kim Yong-beom Reveals Anthropic's Regulatory Discussions
Kim Yong-beom met with reporters at the San Francisco Press Center on the 24th (local time) and stated, "The Ministry of Science and ICT is continuing dialogue with Anthropic, and they explained that they themselves are talking with Washington D.C. regulatory authorities." Kim conveyed that Anthropic explained, "The company has developed the product and wants to make it available to everyone, but government guidelines prevent this - the company's basic principle is that it wants to make it available to customers."
The Korean government is prioritizing government access over private sector access in its negotiations. Kim stated, "The Ministry of Science and ICT is conducting negotiations with top priority so that the government can access it before private companies," adding that "Anthropic also explained that they are continuing dialogue with US regulatory authorities."
Mythos Model Capabilities and US Security Restrictions
Mythos is a next-generation AI model developed by Anthropic. It is known to possess cybersecurity capabilities advanced enough to independently explore system vulnerabilities and design attack scenarios. For this reason, the US government classifies it as strategic technology directly linked to national security and strictly restricts access.
The Korean government is also continuing negotiations with the US side to access Mythos, but general access has not yet been permitted.
NVIDIA Expresses Similar Position on Model Access
NVIDIA conveyed a similar position regarding model access debates. Kim stated, "NVIDIA also made similar comments related to the debate surrounding open models and closed models within the United States," adding that "the company has the position that both types of models are necessary and is making persuasive efforts in that direction within the United States."
FAQ
What did Kim Yong-beom reveal about Anthropic's negotiations on the 24th?
Kim Yong-beom revealed at the San Francisco Press Center on the 24th (local time) that Anthropic is conducting direct talks with Washington D.C. regulatory authorities while the Ministry of Science and ICT continues separate dialogue with the company regarding Mythos model access.
Why does the US government restrict access to the Mythos AI model?
The US government classifies Mythos as strategic technology directly linked to national security because the model possesses advanced cybersecurity capabilities, including the ability to independently explore system vulnerabilities and design attack scenarios, leading to strict access restrictions.
