The U.S. government lifted a previous ban restricting foreign access to Anthropic’s artificial intelligence models, Fable 5 and Mythos 5, allowing the company to resume service to international users. This action reversed a directive that forced Anthropic to disable these models worldwide since mid-June under national security concerns.
The original order barred foreign nationals, including Anthropic employees outside the U.S., from accessing the AI systems. Officials reportedly feared that Fable 5 might have vulnerabilities that allowed it to be “jailbroken,” though extensive testing involving government agencies and private parties had not uncovered any universal security flaws. The recent decision followed a partial easing weeks earlier when Mythos 5 access was granted to select companies and federal agencies under close government oversight.
Commerce Secretary Howard Lutnick communicated that Anthropic no longer needed an export license after the company committed to proactively detecting security risks, collaborating with authorities on future releases, and reporting malicious activities linked to its technology. This move enables the broad restoration of Fable 5 access both domestically and internationally.
The ban originated from concerns about supply chain risks flagged by the Defense Department, which had also restricted the Pentagon’s use of Anthropic’s technology earlier in the year. Anthropic had sought to limit use cases involving fully autonomous weapons and mass domestic surveillance, drawing a line between technological advancement and ethical considerations.
Fable 5 serves as a consumer-facing AI model designed with enhanced safeguards compared to the more powerful Mythos 5, aiming to mitigate misuse in cyberattacks. The decision to lift restrictions responded in part to pressure from AI companies and European governments eager to utilize Anthropic’s tools to identify and patch potential vulnerabilities before hostile actors could exploit them.