By Emilio Crespo
The Trump Administration’s authorization for Anthropic to reopen its next-generation public AI has a catch. The model is not what it was when it was banned in June.
The company that leads the development of Artificial Intelligence in programming and scientific uses has bowed to national security criteria and offers a version that has the same name, Fable 5, but has less capacity and more limitations.
The opinion is widespread among those who have compared it with the original version, open to the public for only two days until its ban for non-American users, something so difficult to apply that in practice it meant its closure. Anthropic has also acknowledged the forced cropping of the creature.
Fable was already a layered version of another AI, Mythos, which is still limited to a few dozen firms approved by the US Government.
OpenAI, creator of the most popular AI, ChatGPT, is immersed in a similar process with its latest model, GPT-5.6, considered as powerful as Mythos and whose access is limited to entities authorized by Washington,
The White House justifies the restrictions by saying that public access to tools capable of manufacturing biological weapons or penetrating government computer systems would be a threat to national security. For those looking for the latest in AI there is no alternative given the enormous advantage of the US over China and Europe, whose best models, the Chinese DeepSeek and the French Mistral, have also opted for open source.
The masses can continue to marvel at less sophisticated AIs and the social revolution they are fueling. But the most powerful models are only for those who decide the US Government. The free market does not govern these technologies.