US bars foreign access to Anthropic's Fable 5 and Mythos 5 AI models
Synopsis
For the first time, the US government has ordered an AI company — Anthropic — to cut off all foreign nationals from its two most powerful models, Fable 5 and Mythos 5, just days after launch. The move treats frontier AI software as a national-security asset, setting a precedent that could reshape how every major AI lab deploys its models globally.
Key Takeaways
Washington ordered Anthropic to block all foreign nationals from Fable 5 and Mythos 5 on Friday, 13 June 2026 at 5:21 pm Eastern time .
The directive is the first US export-control measure to target specific AI models by name.
The restriction applies to Anthropic 's own foreign-born employees, not just external customers.
Fable 5 and Mythos 5 were released on Tuesday, 10 June 2026 and marketed as Anthropic 's most capable models to date.
Mythos 5 had been previewed two months earlier under Project Glasswing , touting unprecedented cybersecurity capabilities.
Access to all other Anthropic models remains unaffected, according to the company.
Washington has directed Anthropic, the Silicon Valley-based AI company, to immediately block all foreign nationals from accessing its two newest frontier models — Fable 5 and Mythos 5 — citing national security concerns, according to the company. The directive, received at 5:21 pm Eastern time on Friday, 13 June 2026, marks the first time the US government has imposed export controls targeting specific AI models by name.
An Unprecedented Export-Control First
The order represents a significant escalation in Washington's approach to regulating frontier artificial intelligence. Unlike prior restrictions focused on hardware — particularly semiconductor exports — this directive goes directly after software models, setting a new precedent for how the Trump administration intends to govern the most capable AI systems. The measure extends even to Anthropic's own foreign-born employees, according to the company's statement.
'The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Anthropic models will not be affected,' the company said.
What Fable 5 and Mythos 5 Are
Anthropic released Fable 5 and Mythos 5 on Tuesday, 10 June 2026, marketing them as its most powerful models to date. Both systems share similar underlying technical foundations, but only Fable 5 was made publicly available — with additional safety and security restrictions in place — while Mythos 5 was released exclusively to trusted partners.
Anthropic had previously warned that Fable 5 'could be misused to cause serious damage' without the safeguards it introduced, including restrictions on discussing sensitive cybersecurity, biology, and chemistry topics.
Project Glasswing and the Mythos Preview
The releases came two months after Anthropic first unveiled a preview version of the Mythos model to a small group of industry partners under an initiative called Project Glasswing, touting unprecedented cybersecurity capabilities. That limited rollout was designed to stress-test the model's most sensitive features before broader deployment.
The speed with which the government moved — intervening within days of the public launch — signals that agencies had been monitoring the release closely and had prepared the directive in advance.
Why It Matters
The export-control order arrives amid intensifying geopolitical competition over advanced AI, particularly between the United States and China. By restricting foreign access at the model level, Washington is effectively treating cutting-edge AI as a strategic asset on par with advanced semiconductors. The move could prompt scrutiny of other frontier AI developers, including those at OpenAI, Google DeepMind, and Meta.
What's Next
It remains unclear whether Anthropic will seek a waiver or appeal the order, or whether the restriction is intended to be permanent. The company has not indicated a timeline for restoring access. Analysts and policymakers will be watching whether US Commerce Secretary Howard Lutnick and the broader Trump administration extend similar controls to other leading AI models, and how allied governments respond to the sudden loss of access for their own researchers and enterprises.
Point of View
Washington is now moving up the stack to restrict the software layer itself. What makes this qualitatively different is that AI models are dual-use in a way chips are not: they can be replicated, fine-tuned, and redeployed with minimal hardware. By targeting Fable 5 and Mythos 5 specifically, the administration is signalling that it views model weights and capabilities as export-controlled munitions, a framing that mainstream coverage has largely underplayed. The inclusion of Anthropic's own foreign-born staff in the restriction is the most consequential detail: it suggests the government is prepared to accept significant operational disruption at a flagship domestic AI lab to enforce the policy, raising immediate questions about talent retention across the entire sector.
NationPress
29 Jul 2026
Frequently Asked Questions
Why did the US block access to Anthropic's Fable 5 and Mythos 5?
The US government cited national security concerns when it ordered Anthropic to disable Fable 5 and Mythos 5 for all foreign nationals on 13 June 2026. The directive is the first export-control measure to target specific AI models by name, reflecting Washington's view that frontier AI capabilities pose strategic risks if accessed abroad.
Who is affected by the Anthropic foreign-access ban?
All foreign nationals are barred from accessing Fable 5 and Mythos 5, including Anthropic's own foreign-born employees, according to the company's statement. External customers outside the US are also cut off, though access to all other Anthropic models remains unaffected.
What are Fable 5 and Mythos 5?
Fable 5 and Mythos 5 are Anthropic's most powerful AI models to date, released on Tuesday, 10 June 2026. Both share similar technical foundations; Fable 5 was released publicly with safety restrictions, while Mythos 5 was made available only to trusted partners and had been previewed two months earlier under Project Glasswing.
What is Project Glasswing?
Project Glasswing was an Anthropic initiative that gave a small group of industry partners early access to a preview version of the Mythos model approximately two months before its full release. The programme highlighted the model's unprecedented cybersecurity capabilities and served as a controlled introduction before broader deployment.
Could other AI companies face similar export restrictions?
The Anthropic order sets a direct precedent that could expose other frontier AI developers — including OpenAI, Google DeepMind, and Meta — to comparable directives for their most capable models. Analysts will be watching whether the Trump administration and US Commerce Secretary Howard Lutnick broaden the policy to additional AI systems in the coming months.
