US government orders Anthropic to block AI models for foreign users over security concerns

Source Domain: mezha.net

Sudden Government Directive: The U.S. government issued an urgent export control directive on June 13, requiring the termination of access to the Fable 5 and Mythos 5 AI models for all foreign nationals, including within the country and foreign employees of Anthropic.
Lack of Clear Justification: The reason behind the directive remains unclear, raising concerns across the AI industry about the transparency of government actions regarding AI models and their safeguards.
Potential Security Risks: Anthropic suggested that the directive may stem from a single demonstrated bypass method revealing minor vulnerabilities, but these risks were already mitigated by their extensive safeguards.
Compliance Despite Disagreements: Anthropic agreed to disable access to the affected AI models, despite disagreeing with the directive and expressing its belief in the necessity for fair, transparent, and evidence-based regulation of AI deployment.
Call for Transparent Regulation: Anthropic is advocating for clear, transparent, and technically sound processes for regulating AI models, emphasizing that the government should have the capability to block dangerous deployments, but only through fair and evidence-based procedures.

US government orders Anthropic to block AI models for foreign users over security concerns | Ukraine news