The US government has finally given a green signal to Anthropic to restore limited access to its most advanced Mythos 5 AI model after the company addressed concerns related to potential national security.
This decision comes just two weeks after the Trump administration ordered the Dario Amodei-led firm to stop giving foreign nationals access to its Mythos 5 and the Fable 5 model. The move had effectively shut down global access to both advanced AI tools and sparked widespread debate across the industry on who controls AI and who is allowed to not.
According to Bloomberg, the latest approval signals a partial breakthrough.
What Has Changed?
Reportedly, the US administration has approved the return of Mythos 5 to a limited group of trusted cybersecurity organisations and infrastructure providers. US Commerce Secretary Howard Lutnick reportedly mentioned that the company had made significant progress in addressing the government’s concerns about risks linked to the model. As a result, Mythos 5 can once again be shared with selected partners.
However, the restrictions on Fable 5 remain in place. It means that the broader public still cannot use the Fable 5 AI tool.
The US government feared that users could bypass or jailbreak the model’s built-in safety features. Moreover, the officials were particularly concerned about the Mythos 5’s abilities to identify critical software flaws. If misused, such technology may potentially help attackers discover weakness in computer systems, especially by Russia and China, noted the Trump government.
This latest decision is crucial far beyond the United States. Although Anthropic is an American firm, its AI tools are used by researchers and firms across the world. The earlier restrictions had affected international users because access to Mythos 5 was suspended globally.
To recall, a legal tech firm called Legion LegalTech Corp, has filed a lawsuit earlier this week against the US government, accusing that the order given by the administration banning the Mythos 5 and Fable 5 has disrupted its operations and affected employees working outside the US.
Notably, the new approval means that some trusted international partners may regain access, but many organisations outside the US may still face ban until the government expands the list of approved users. As per Reuters, more than 100 companies and institutions will now have access to Mythos 5, comprising multiple Fortune 500 firms.
This case shows that access to powerful AI models could depend on government approval rather than company decisions alone. Countries developing their own AI policies will likely watch this incident as they decide how to innovate with AI and maintain their stance on its usage for the rest of the world. Moreover, the future of advanced AI will depend not only on technological progress, but also on security, regulation and international trust.
