The United States government has removed export restrictions on Anthropic’s most advanced artificial intelligence models, Claude Fable 5 and Mythos 5, allowing the company to restore international access after weeks of limited availability. The move marks a significant shift in the US administration’s approach to regulating cutting-edge AI technologies and comes amid growing debate over balancing national security with innovation.
Anthropic to Restore Global Access
Anthropic announced late Tuesday that it would begin restoring access to Claude Fable 5 and Mythos 5 starting the following day after receiving official notification from the US Department of Commerce that export controls had been lifted.
In a statement shared on X, the San Francisco-based AI company thanked its users for their patience during the disruption and acknowledged those who worked on redeploying the models. The announcement ends weeks of uncertainty for international customers and researchers who had been locked out of the company’s most powerful AI systems.
Government Sets New Security Conditions
According to a letter from US Commerce Secretary Howard Lutnick, widely circulated online, Anthropic will no longer require export licences to provide access to the advanced models. The decision follows the company’s agreement to adopt stricter security measures in collaboration with the US government.
Under the new arrangement, Anthropic has committed to proactively identifying and mitigating security risks associated with its AI models, cooperating with federal authorities in developing safety standards for future systems, and reporting instances of malicious activity involving its technology.
These commitments were viewed as key factors in convincing regulators that broader international access could resume without compromising national security.
Restrictions Sparked by Security Concerns
The restrictions were imposed last month after the administration of US President Donald Trump directed Anthropic to prevent all foreign nationals—including some of its own employees—from accessing Claude Fable 5 and Mythos 5.
At the time, Anthropic said the government had not provided a detailed explanation beyond citing national security concerns. However, the company believed officials were particularly worried about potential security vulnerabilities in the Fable 5 model following reports suggesting it could be “jailbroken” to bypass built-in safeguards.
Last week, the company received approval to provide the models to US organisations responsible for operating and protecting critical infrastructure while negotiations continued to restore broader public access.
AI Oversight Continues to Intensify
The temporary restrictions highlighted ongoing tensions between Anthropic and the Trump administration. Earlier this year, the company sued the US Department of Defense after being designated a “supply chain risk” because it declined to work with the military without guarantees that its AI systems would not be used for mass surveillance or autonomous weapons.
Although the Trump administration initially promoted a lighter regulatory approach to artificial intelligence, recent developments suggest increased scrutiny of advanced AI technologies. OpenAI, for instance, recently announced that its upcoming GPT-5.6 models would initially be released only to a limited group of trusted partners following government pressure to adopt a phased rollout.
Commenting on the latest decision, Francesco Bailo, Deputy Director of the AI, Trust and Governance Centre at the University of Sydney, said the reversal had been widely anticipated within the technology community. He noted that concerns surrounding the reported vulnerabilities in Fable 5 had likely been overstated and that maintaining strict export controls could have created an undesirable regulatory precedent while drawing strong opposition from the AI industry.
Also Read :- Anthropic Faces Growing Questions Over AI Spending Ahead of IPO
