Anthropic reactivated its most advanced AI model for public access Wednesday, hours after the Trump administration reversed export restrictions that had blocked the system from broad release.
Fable 5 represents the cutting edge of publicly available artificial intelligence, powerful enough to trigger government concern. The restrictions came with a requirement that Anthropic implement additional safeguards before allowing widespread use.
All customers now have access to Fable, though Anthropic will automatically reroute queries flagged as security or safety risks to less capable models. The company is also managing availability through pricing mechanisms. Most users will need to pay per token consumed rather than draw from a subscription package.
Through July 7, subscribers can use up to half their monthly token allocation on Fable, though the company warns the model consumes tokens far faster than its other offerings. After that date, the subscription benefit expires.
The episode has highlighted broader uncertainty about government intervention in frontier AI development. OpenAI is currently withholding the broad release of GPT 5.6 after the government requested a delay, signaling a pattern where the administration may selectively constrain the availability of cutting-edge systems.
The question now facing the industry is what threshold triggers intervention and which companies will face similar restrictions. Anthropic's experience suggests the administration is willing to use export controls as a lever, but the consistency and scope of future actions remain unclear.
Author James Rodriguez: "The Trump administration just proved it can flip the switch on AI releases overnight, and nobody knows what comes next."
Comments