Anthropic says it will begin restoring access to Claude Fable 5, bringing back a model that had been sidelined for weeks after the US government imposed emergency export controls. The Verge reports that Anthropic announced on X that the Department of Commerce had lifted export controls on both Claude Fable 5 and Mythos 5, with restoration set to begin the next day.
That makes this more than just a product comeback. Fable 5 had become one of the clearest examples yet of how fast-moving AI development can suddenly collide with national security policy. Reuters reports that the Commerce Department’s earlier order had forced Anthropic to disable both models after concerns they could be exploited by adversaries, including through possible jailbreaks.
What Happened to Fable 5
According to The Verge, Anthropic pulled Fable 5 in early June after what it described as a Friday evening ultimatum from the Trump administration. The restrictions did not just apply to foreign customers overseas. The directive barred access for foreign nationals, including non-US staff at client companies and even many of Anthropic’s own employees.
The company had described Fable 5 as its consumer-facing model built on the same underlying technology as Mythos 5, but with more safeguards. That meant the export action hit both Anthropic’s most advanced internal-class system and the public-facing variant built around it.
The US Has Now Reversed Course
Reuters says the US has now fully withdrawn those export controls, meaning licenses are no longer required. The report adds that Anthropic pledged to detect and address security risks, coordinate with US authorities on model protocols, and report malicious activity.
That is a significant shift from just a few days earlier. The Verge notes that Mythos 5 had already been allowed back in a more limited form for a pre-approved list of organizations, while Fable 5 remained out of reach. Now, with the broader controls lifted, Anthropic can start bringing both models back.
Why This Story Matters Beyond Anthropic
The bigger issue here is not only whether Fable 5 is back. It is what this episode says about the future of frontier AI regulation.
This was one of the clearest signs yet that governments are willing to intervene directly when they think a model’s capabilities pose national security risks. Reuters reports that OpenAI’s GPT-5.6 also faced a restricted, staggered debut under similar rules, initially limited to vetted partners and government departments.
That suggests the fight is no longer just about who has the strongest model. It is also about who gets to access those systems, under what rules, and how quickly governments will step in when they think the risks are too high. Anthropic’s clash with Washington may have looked like one company’s problem, but it increasingly feels like a preview of what advanced AI rollouts could look like from here.
Anthropic Still Has a Bigger Problem to Solve
Even with Fable 5 returning, this was not a clean win for Anthropic. The Verge notes that the company has already been feuding with the government over a separate supply chain risk designation while also preparing for an IPO. In other words, the model is back, but the broader standoff between Anthropic and Washington is not necessarily over.
That makes Fable 5’s return feel less like the end of a controversy and more like the next chapter in a bigger argument over how powerful AI systems should be controlled, who should be trusted with them, and how governments decide where the red lines are.
Why this matters for Australia
Australian users, developers and businesses are not sitting outside this story. When the US throws export controls around a frontier AI model, the effects can spill across international teams, global enterprise customers and non-US workers almost immediately. Reuters and earlier reporting on the original order showed just how broad those restrictions were.
It also highlights something bigger for Australian readers: access to the most advanced AI models may increasingly depend not just on what a company builds, but on what governments allow. That makes AI policy feel less abstract and much more tied to who can actually use the tools.
The broader takeaway is simple: Claude Fable 5 is back, but the episode shows that frontier AI access can now be turned on and off by geopolitics as much as by product roadmaps.
Source links
Source: The Verge | Reuters
