US Government Blocks Anthropic Models: AI Control Tightens

The US Department of Commerce has unexpectedly imposed severe restrictions on Anthropic, forcing the company to take its most advanced AI models offline. This decision caused a major stir in the tech world, as it demonstrated the government's ability to directly intervene in the products of private IT companies without a court order. This incident is being interpreted as a serious warning not only for Anthropic but for the entire AI industry. This is reported by Techcrunch.com reports says.

Last Friday, the Department of Commerce sent an official letter to Anthropic, relying on a rarely used export control directive. It prohibited non-US citizens, including the company's foreign employees, from using the Fable 5 and Mythos 5 models. The government explained this as a threat to national security, though it did not disclose specific details. As a result, the company was forced to completely shut down its most powerful models for all users to avoid breaking the law.

Political pressure or technical error?

Initial assumptions suggested that the restrictions were related to the possibility of bypassing the AI's safety systems (guardrails). However, according to Axios, the situation is driven more by political factors than technical issues. Reports indicate that personal disagreements between Anthropic leadership and the Donald Trump administration may have triggered the adoption of the export directive.

Renowned cybersecurity expert Katie Moussouris analyzed the situation and questioned the government's claims. According to her, the security flaws identified by Amazon researchers were not serious enough to warrant the application of export controls. Moussouris emphasized that the models are critical for cyber defense and blocking them weakens US network defenders.

Experts believe the difference between asking an AI model to check code for security and asking it to fix it is very small. The government, however, found exactly such functions to be dangerous. Moussouris and dozens of other specialists have called on the Trump administration to reverse this decision, stating that such steps damage defense capabilities in the field of cybersecurity.

This event is also an important signal for developing markets like Uzbekistan. The fact that global tech giants can restrict their products at any time under US government pressure raises questions about the stability of AI-based services. This once again proves how important it is to develop localized and independent technological solutions.

For now, it remains unknown when the Anthropic models will be back online. This situation indicates that state control over AI in the US has entered a new stage. From now on, tech companies will have to seriously account for not only innovation but also the political climate.