Updated 8:23 AM PDT, July 1, 2026
SAN FRANCISCO — The Trump administration has lifted the ban on new versions of Anthropic’s Claude chatbot, which had been in place due to concerns related to cybersecurity.
On Tuesday night, Anthropic announced that its AI model, Claude Fable 5, is now widely accessible. It is also reinstating access to its more advanced model, Mythos 5, but only for a limited number of U.S.-based organizations that have received federal approval.
The Commerce Department had prohibited foreign nationals from using both models starting June 12, leading Anthropic to temporarily withdraw the products just days after their introduction.
In a blog post this week, Anthropic explained that federal concerns arose from findings by cybersecurity researchers at Amazon, its primary cloud service provider. According to the company, Amazon discovered a way to bypass Fable 5’s security measures, potentially exposing vulnerabilities.
Officials have become increasingly apprehensive since Anthropic indicated earlier this year that its Mythos model was skilled at identifying software flaws that could be exploited by malicious actors, putting critical computer networks at risk globally.
Meanwhile, Anthropic’s main competitor, OpenAI, revealed that it is also delaying the release of its new AI model at the insistence of the Trump administration. OpenAI’s upcoming product, GPT-5.6 Sol, will be made available only to a select group of government-sanctioned customers for a limited time.
Last month, Trump signed an executive order aimed at AI oversight that created a framework for the federal government to assess national security risks posed by advanced AI systems prior to their public release, with a review period lasting up to 30 days. Although participation from AI developers is described as voluntary, the framework has yet to be fully established.
