Anthropic's Fable 5 is available again after Trump's export controls are lifted

🕒 Published on Zendoric: July 3, 2026 · 01:20
Anthropic's Fable 5 model became operational again for users on Wednesday, after the Trump administration lifted on Tuesday night an export control that had kept it blocked.
Anthropic's Fable 5 model is back up and running for users this Wednesday, after the Trump administration lifted on Tuesday night an export control that had kept it blocked. According to Axios, it is the most powerful public AI tool currently available, so capable that the US government itself had required Anthropic to add additional safety measures before allowing it to be broadly available.
Anthropic has confirmed that Fable is now accessible to all customers, although queries the company deems to pose safety risks may be redirected to less powerful models as a precautionary measure. An important nuance is the access model: most customers who want to use Fable will have to do so outside their usual subscription plans, paying directly for the tokens consumed.
As a transition period, Anthropic has announced that until July 7 subscribers will be able to use Fable up to half of the data quota included in their plans, though it warns that the model consumes tokens faster than other alternatives, which in practice limits extended use even within that grace margin.
The background to this episode is especially relevant: the temporary blocking of Fable by the US government has opened a significant debate about how and when the Trump administration might intervene in future frontier-model launches. This precedent is not isolated: OpenAI has stated that it is holding back the broad launch of its latest model, GPT 5.6, while it holds consultations with the government, which expressly requested that delay.
This case illustrates an emerging trend in which US regulators are taking a more active and direct role in controlling the public availability of the most advanced AI models, establishing a de facto mechanism of government review before the commercial deployment of technologies deemed to be of the highest capability. For the agentic AI ecosystem, this raises questions about the predictability of access to the most powerful tools and about how companies will have to navigate this new layer of regulatory oversight in their product launch cycles.
🔗 Related on Zendoric
- The deal Anthropic is negotiating with Washington reopens an uncomfortable question: how do you control the export of something that travels in a file · 2026-06-27
- Washington Reverses Course on Anthropic Export Curbs — And the Real Question Isn't Who Won · 2026-07-01
- Mythos and Fable: when the U.S. applies 20th-century export controls to 21st-century AI models · 2026-06-28


