Anthropic, a prominent AI firm, has announced that it will begin restoring access to its advanced models, Claude Fable 5 and Mythos 5, following the lifting of export controls by the U.S. government. In a statement released on social media platform X, Anthropic expressed gratitude to its users for their patience and to those involved in redeploying the models.
The U.S. Department of Commerce has officially informed Anthropic that it no longer requires the company to obtain an export license for these models. This decision is based on the company’s commitment to proactively identify and mitigate security risks, collaborate with the government on developing standards for future models, and promptly report any malicious activities.
Last month, Anthropic had to abruptly halt access to Claude Fable 5 and Mythos 5 after an order from the Trump administration barred all foreign nationals, including company employees, from interacting with these models. The government cited unspecified national security concerns as the reason for this drastic measure, with particular worries about potential security vulnerabilities linked to Fable 5.
Following the recent developments, Anthropic has received authorization to provide these models to U.S. organizations that are responsible for protecting critical infrastructure and is working to restore general public access.
The previous restrictions had strained relations between Anthropic and the Trump administration. Earlier this year, the company filed a lawsuit against the Department of Defense after being classified as a “supply chain risk” due to its reluctance to collaborate with the military without assurances that its AI technologies would not be utilized for mass surveillance or autonomous armament systems.
Despite the administration initially embracing a more hands-off approach to AI regulation during Trump’s second term, recent actions indicate a shift towards increased oversight of the blossoming technology.
Meanwhile, OpenAI, the organization behind ChatGPT, recently announced it would release its new AI model series, GPT-5.6, initially to a limited group of trusted partners, a response to mounting pressure from the U.S. government to stagger AI releases.
Experts observing the situation view the lifting of restrictions as a logical move. Francesco Bailo, deputy director at the AI, Trust and Governance Centre at the University of Sydney, suggested that the government likely recognized an overreaction on its part and the potential for dangerous precedents in regulatory frameworks.
Similarly, Tanishq Abraham, formerly of Stability AI and now leading the medical AI firm Sophont, noted that the latest developments indicate an improving relationship between Anthropic and the U.S. government, attributing this progress to efforts by co-founder Tom Brown. However, he raised critical questions about the regulatory landscape for frontier AI models, particularly regarding government involvement in future releases.
As the tech industry navigates these complexities, industry stakeholders will be closely watching how the situation unfolds and what implications it may hold for the future of AI governance.
