US Restricts Access to Anthropic’s Powerful AI Models Over Security Concerns
Brief By Newsbrief / 10:20 AM on 16 Jun 2026
The United States has imposed new restrictions on two advanced AI models developed by AI research company Anthropic, known as Fable 5 and Mythos 5. Under the new export control directives, access to these models will be restricted for non-U.S. citizens both inside and outside the United States.
According to media reports, concerns surrounding China played a significant role in the decision. U.S. officials reportedly feared that if groups linked to China gained access to these powerful AI systems, they could pose risks to American national security. However, there is currently no official confirmation that any Chinese organization had obtained access to the models.
Fable 5 and Mythos 5 are among the first publicly released AI systems featuring capabilities associated with Anthropic’s highly advanced Mythos-level technology. Reports suggest that the Mythos AI framework is capable of identifying software vulnerabilities with remarkable efficiency, uncovering weaknesses that human researchers might take months to detect.
Although Anthropic had implemented safety mechanisms designed to prevent the models from providing sensitive cybersecurity-related information, a report claimed that researchers from Amazon successfully bypassed some of these safeguards. Using specialized prompts, they reportedly managed to extract information related to software vulnerabilities, raising concerns about the models’ potential misuse.
The White House and U.S. security agencies are particularly worried that advanced AI systems could be exploited for cyber operations or replicated through AI distillation techniques. Officials fear that foreign adversaries could use such methods to copy key capabilities without directly accessing the original models.
In response, Anthropic has been engaged in discussions with the U.S. government regarding the new regulations. Engineers and senior executives from the company are reportedly working with federal authorities to find a solution that balances national security concerns with continued AI innovation and research.
The restrictions highlight the growing geopolitical competition surrounding advanced artificial intelligence technologies and their potential impact on global security.