Anthropic abruptly disabled access to its advanced AI models Fable 5 and Mythos 5 for all foreign users after the US government imposed a sweeping export control order, highlighting growing tensions around AI governance and international security.
- US Commerce Department restricts foreign access to Anthropic's AI models
- Security concerns arise from demonstrated jailbreaks and potential cyber threats
- EU allies criticize export ban as discriminatory and disruptive
What happened
Anthropic PBC was forced to disable foreign access to its two most powerful AI models, Fable 5 and Mythos 5, shortly after their launch. The action followed a US Commerce Department export control order that explicitly banned all foreign nationals, including those from allied countries and Anthropic's own foreign employees, from accessing these models.
The export ban was triggered after national security authorities expressed urgent concerns regarding vulnerabilities in the models that could be exploited for malicious purposes. UK researchers from the AI Safety Institute announced significant progress in developing a jailbreak for Fable 5, while Amazon researchers warned the US government about potential cyberattack-related risks revealed through their testing.
Why it matters
This move underscores how AI technology is becoming a focal point of geopolitical tension, with governments increasingly viewing frontier AI capabilities through a national security lens. By restricting foreign access to advanced AI models, the US is asserting control over sensitive AI technologies and aiming to mitigate risks related to cybersecurity and potential misuse.
The export restriction disrupts Anthropic's international collaborations and impacts its global workforce, who are now barred from using the very models they helped develop. Furthermore, the ban has sparked criticism from European allies who fear that such unilateral restrictions may harm research, innovation, and international cooperation in AI development.
What to watch next
Attention will focus on how Anthropic and other AI labs respond to increasing regulatory pressures and whether new frameworks for AI export control will emerge. The ongoing debate around the balance between innovation, security, and international cooperation will continue to shape AI governance globally.
European Union officials are currently assessing the implications of the US ban and may push for diplomatic or regulatory responses. Industry stakeholders will also watch closely for any changes to deployment strategies by AI developers and how countries align their policies to address cybersecurity risks without stifling technological progress.