Anthropic’s Claude Mythos 5 and Fable 5 were introduced to the world earlier this month, but as soon as the models were launched, the US government put a ban on the use of the AI models. Now, after weeks of conversations with the White House, Anthropic has received partial approval for restoring access to Mythos 5 for a select group of trusted organisations from the US government. However, the company’s plans to make Fable 5 widely available remain on hold. Government officials said the change followed progress in talks with Anthropic on reducing security risks linked to advanced AI systems. Although the move restores access for some users, most of the restrictions introduced earlier this month continue to remain in effect until further notice. Here is the full story in 5 simple points:
Anthropic officially launched Claude Mythos 5 and Claude Fable 5 on June 9, 2026. Mythos 5 is the more powerful model and is available only to selected organisations. It was designed to handle complex reasoning, cybersecurity tasks, and critical infrastructure management.
Whereas the Claude Fable 5 was only available to enterprise users across the globe. Much like Mythos 5, the Fable 5 were also capable of handling complex reasoning, cybersecurity tasks, and critical infrastructure management. However, the main difference was that it was built for broader enterprise use, and the capabilities were controlled and restricted when compared to those of the Claude Mythos 5.
The US government restricted access to both models on June 12, 2026. Officials said that they were concerned that the advanced capabilities of the AI systems could be used to identify zero-day software vulnerabilities or modified for malicious purposes.
Furthermore, the administration also classified the models as ‘dual-use strategic goods’ and argued that scammers, hackers and notorious people could potentially misuse them to target important American infrastructure.
Also read: Your next iPhone may cost more: Here is why an iPhone price hike likely on cards
After the restrictions were imposed, Anthropic entered into the discussion mode with the US Department of Commerce. Commerce Secretary Howard Lutnick, in a letter on June 26, said that the company agreed to add more detailed safety measures and stronger access controls.
He further added that the talks led to a revised compliance framework that satisfied government concerns while allowing limited use of the model in high-security environments and allowed the company to restore access to Mythos 5 for certain organisations. However, the future of the Fable 5 is still uncertain, and both parties are still discussing it.
Anthropic started restoring access to around 100 pre-approved trusted partners as soon as they received the go-ahead from the US government. It includes selected federal agencies and critical infrastructure providers as well.
The updated policy also allows these organisations to use non-US employees within approved environments, reversing parts of the earlier blanket restriction. However, Fable 5 remains unavailable and has not been included in the latest restoration.
Also read: OpenAI unveils GPT 5.6 family of AI models, but you can’t use them yet: Here is why
While the access has been granted to certain US-based companies, there is no certainty if and when the access will be provided to the Indian firms or any other firms in the world. This has led many to believe that the frontier AI models are increasingly being treated as strategic infrastructure rather than ordinary software.
Furthermore, it also highlights a growing reality that access to advanced AI may depend on government policies as much as technology providers. Not only that but it also raises concerns about operational continuity, encouraging organisations to diversify across multiple AI platforms instead of relying on a single model that could face sudden regulatory restrictions.