On Friday 13 June, Anthropic was ordered by the US government to pull Fable 5 and Mythos 5 offline. Both models had been publicly available for roughly 72 hours.
As of today, they’re still down, with no confirmed timeline for return.
Here’s what we know.
What Happened
At 5:21 PM US Eastern Time on Friday, Anthropic received an export control directive from the US Commerce Department. The order required Anthropic to immediately suspend access to Fable 5 and Mythos 5 for all foreign nationals, including those based inside the United States, and including Anthropic’s own foreign national staff.
Because Anthropic said it had no reliable way to verify citizenship status in real time, it took the only path it could see to ensure compliance: it disabled both models for every customer globally.
All other Claude models remain available, including Opus, Sonnet and Haiku.
Why the Government Stepped In
The official reasoning has been thin. Anthropic says the directive arrived with no detailed explanation of the underlying national security rationale.
What has surfaced publicly is that US officials were concerned about a potential jailbreak, specifically a method to prompt Fable 5 to identify vulnerabilities in a codebase. Anthropic’s position is that this represents a “narrow, non-universal” capability, and one that’s already available in other widely used models including OpenAI’s GPT-5.5.
Anthropic has described the situation as a misunderstanding and said it’s working to restore access as quickly as possible. In a public statement, the company also warned that if this standard were applied broadly, it would effectively halt all new frontier model deployments across the industry.
Whether you read that as a principled defence of open AI access or a company protecting its commercial interests ahead of a planned IPO, both things can be true at once.
What This Means If Your Business Uses Claude
The short version: if your teams use Claude directly through Anthropic’s platform, they will be running on Opus 4.8 or another available model until Fable 5 is restored. The day-to-day experience should be largely unchanged for most users, given Fable 5 had only been live for a few days before the suspension.
If your teams access Claude through Microsoft 365 Copilot Chat, there’s no disruption to that integration. The Claude models surfacing inside Microsoft 365 are from the standard tier and are not affected by this directive.
The Bigger Question This Raises
This is where I think the more useful conversation sits, and it’s one that applies well beyond Anthropic.
What happened last week is a reminder that when your business relies on a third-party AI model, access to that capability can disappear overnight. Not because the vendor made a bad product decision. Not because of a security incident on your side. But because a government regulator issued a directive at 5:21 on a Friday afternoon.
That’s not a reason to avoid using AI. It’s a reason to think carefully about which capabilities are genuinely core to your business versus nice to have, and whether you have a fallback when a model changes, gets deprecated or, as in this case, gets pulled without warning.
We’ve seen this pattern emerge before with Copilot agent sprawl and the risk of building workflows around AI features that can shift with a licensing update. The Fable 5 situation is a more dramatic version of the same underlying challenge: AI vendors are still operating in a regulatory environment that’s catching up with the technology, which means the ground can shift quickly.
A practical AI operating model doesn’t just help you choose the right tools today. It helps your business stay productive when those tools change tomorrow.
Where Things Stand Right Now
Fable 5 and Mythos 5 are offline globally with no confirmed return date.
Anthropic is contesting the directive and says it believes the order is based on a misunderstanding.
All other Claude models are unaffected.
Claude within Microsoft 365 Copilot Chat continues to operate normally.
Anthropic has not updated its public statement since Friday.
I’ll update this post when there’s a confirmed change.
About the Author
Carlos Garcia is the Founder and Managing Director of CG TECH, where he leads enterprise digital transformation projects across Australia.
With deep experience in business process automation, Microsoft 365, and AI-powered workplace solutions, Carlos has helped businesses in government, healthcare, and enterprise sectors streamline workflows and improve efficiency.
He holds Microsoft certifications in Power Platform and Azure and regularly shares practical guidance on Copilot readiness, data strategy, and AI adoption.
On Friday 13 June, Anthropic was ordered by the US government to pull Fable 5 and Mythos 5 offline. Both models had been publicly available for roughly 72 hours.
As of today, they’re still down, with no confirmed timeline for return.
Here’s what we know.
What Happened
At 5:21 PM US Eastern Time on Friday, Anthropic received an export control directive from the US Commerce Department. The order required Anthropic to immediately suspend access to Fable 5 and Mythos 5 for all foreign nationals, including those based inside the United States, and including Anthropic’s own foreign national staff.
Because Anthropic said it had no reliable way to verify citizenship status in real time, it took the only path it could see to ensure compliance: it disabled both models for every customer globally.
All other Claude models remain available, including Opus, Sonnet and Haiku.
Why the Government Stepped In
The official reasoning has been thin. Anthropic says the directive arrived with no detailed explanation of the underlying national security rationale.
What has surfaced publicly is that US officials were concerned about a potential jailbreak, specifically a method to prompt Fable 5 to identify vulnerabilities in a codebase. Anthropic’s position is that this represents a “narrow, non-universal” capability, and one that’s already available in other widely used models including OpenAI’s GPT-5.5.
Anthropic has described the situation as a misunderstanding and said it’s working to restore access as quickly as possible. In a public statement, the company also warned that if this standard were applied broadly, it would effectively halt all new frontier model deployments across the industry.
Whether you read that as a principled defence of open AI access or a company protecting its commercial interests ahead of a planned IPO, both things can be true at once.
What This Means If Your Business Uses Claude
The short version: if your teams use Claude directly through Anthropic’s platform, they will be running on Opus 4.8 or another available model until Fable 5 is restored. The day-to-day experience should be largely unchanged for most users, given Fable 5 had only been live for a few days before the suspension.
If your teams access Claude through Microsoft 365 Copilot Chat, there’s no disruption to that integration. The Claude models surfacing inside Microsoft 365 are from the standard tier and are not affected by this directive.
The Bigger Question This Raises
This is where I think the more useful conversation sits, and it’s one that applies well beyond Anthropic.
What happened last week is a reminder that when your business relies on a third-party AI model, access to that capability can disappear overnight. Not because the vendor made a bad product decision. Not because of a security incident on your side. But because a government regulator issued a directive at 5:21 on a Friday afternoon.
That’s not a reason to avoid using AI. It’s a reason to think carefully about which capabilities are genuinely core to your business versus nice to have, and whether you have a fallback when a model changes, gets deprecated or, as in this case, gets pulled without warning.
We’ve seen this pattern emerge before with Copilot agent sprawl and the risk of building workflows around AI features that can shift with a licensing update. The Fable 5 situation is a more dramatic version of the same underlying challenge: AI vendors are still operating in a regulatory environment that’s catching up with the technology, which means the ground can shift quickly.
A practical AI operating model doesn’t just help you choose the right tools today. It helps your business stay productive when those tools change tomorrow.
Where Things Stand Right Now
I’ll update this post when there’s a confirmed change.
About the Author
Carlos Garcia is the Founder and Managing Director of CG TECH, where he leads enterprise digital transformation projects across Australia.
With deep experience in business process automation, Microsoft 365, and AI-powered workplace solutions, Carlos has helped businesses in government, healthcare, and enterprise sectors streamline workflows and improve efficiency.
He holds Microsoft certifications in Power Platform and Azure and regularly shares practical guidance on Copilot readiness, data strategy, and AI adoption.
Sources
Recent Posts
Popular Categories
Archives